This is an old-fashioned weblog, in that it is a list of links I've been reading, with excerpts and notes to myself - sometimes I have lost the original reference. It covers readings around genetics, physics, epidemiology, statistics, programming, philosophy, SF, bushwalking and rockclimbing. It is a cutdown version of an internal weblog, and is updated spasmodically. Additions to earlier topics are often made to the original text.
A minor problem with the EGGX graphics library is the current inability to rotate text with the drawstr command. One way to get around this was to move to using the public domain Hershey vector fonts, which can be easily arbitrarily transformed.
At the moment, the "Times Roman" (now usually called Hershey Serif) fonts from Kamal Mostafa's repository in the original ".jhf" format can now be read into Sib-pair Scheme as a hashed list and written to the graphical window using a command draw-hershey-str.
I read that these are stroke or engraving fonts (as opposed to outline fonts). Windell Oskay presents a number of open licenced stroke fonts, including an extended Hershey font, as SVG font files usable in Inkscape. These are easily converted, as they specify each glyph using just:
| M/m <x> <y> | Move to absolute/relative coordinates x, y |
| L/l <x> <y> | Line to absolute/relative coordinates x, y |
| C/c <x0> <y0> <x1> <y1><x> <y>... | Cubic Bezier curve to absolute/relative coordinates x, y, given control points x0,y0,x1,y1 |
| Z | Line to start of current subpath (ie target of last moveto) |
with coordinates in a 1000 "units-per-em" cell and a horizontal advance of 290-900 units. Here is the EMS Readability font, which I converted to be used by the new (draw-strokes-ch) and (draw-strokes-str) procedures.
(define EMSReadability (hash-table 32 (list "space" (list 0 378) (list "M" 200 378)) 33 (list "exclam" (list 0 280) (list "M" 195 674 "L" 195 192 "M" 173 69.3 "L" 173 18.9 "L" 220 18.9 "L" 220 69.3 "L" 173 69.3 )) 34 (list "quotedbl" (list 0 381) (list "M" 117 693 "L" 120 491 "M" 255 693 "L" 261 491 )) [...] ))
The Relief SingleLine typeface is an open licensed SVG font is another recent alternative to the Hershey fonts, that is now usable from Scheme:
Another such set of fonts is the old GIMM fontset used by the EDWIN drawing package.
These can be converted to other formats using the monobit package, which targets old bitmap graphics formats and a few stroke font formats. The monobit native ".yaff" output format is human readable, and includes the drawing path using relative movement commands.
I used an awk script to convert these to the same Scheme format read by the draw-strokes commands:
| 
|
a subgenre of progressive metal, termed for an onomatopoeia of the guitar sound [...] progressive, rhythmic, and technical complexity accompanied by a use of polymetric groove [...] heavily-distorted, palm-muted guitar chords, syncopated riffs
Core bands: Animals as Leaders, Periphery, Tesseract, Monuments and Textures
Influenced by: A Life Once Lost, Veil of Maya, Vildhjarta, Xerath, and Uneven Structure. Born of Osiris, Scale the Summit
I rely mainly on R base graphics for publication quality data visualizations. R includes "demo(graphics)" which I thought would be a good test for the Scheme plotting procedures I have written. I haven't rewritten these to use other fonts yet, and the default font is not that attractive.
| 
| 
|
| 
| 
|
| 
| 
|
| 
| 
|
Given the EGGX graphics primitives are pretty generic, this statistical plotting library should be reasonably portable.
Still reading (most of these appear in "most popular" lists):
Cultivation Nerd Cultivation Online Lighting Lanterns in the Mountains and Seas Martial God Asura Metaworld Chronicles Millennial Mage Radiant Blade of the Wilderness [the author of Lord of Mysteries] Shadow Slave Star Odyssey Super Supportive The Duke's Son [the author of Overgeared] The Mirror Legacy The Support Ate It All [slowly progressing, but usually amusing] Wait, How Did My Digital Girlfriend Become a Sword Immortal? Zenith of Sorcery [the author of Mother of Learning]
Since I encountered a segfault calling the EGGX makecolor function from Fortran, I wrote a replacement colour mapping function that includes the matplotlib maps such as Viridiana and Inferno. And then a marching squares contouring function. Here is a plot of the Rosenbrock function including the path the varmet minimizer followed to the solution from its start at (-2,-2):
(define x (rseq -2 2 100))
(define y (rseq -2 2 100))
; Just like R's expand.grid
(define xy (expand-grid x y))
(define z (matrix (map (lambda (x y) (rosenbrock (list y x))) (car xy) (cadr xy))
'rows (length x)))
(define line10 (one-contour x y (transpose z) 10))
(define line3 (one-contour x y (transpose z) 3))
(define line1 (one-contour x y (transpose z) 1))
(define fig2 (create-plot '(-2 2) '(-2 2) 500))
(add-matrix-points fig2 "viridiana" x y z 3)
(add-points fig2 '(-2) '(-2) "red" 5)
(add-text fig2 -2 -2 "Start" 20)
;
; This was cut and pasted from the verbose output of varmet
;
(add-points fig2 (map car maxpath) (map cadr maxpath) "red" 2)
;
; I originally rescaled the colour map on a log scale,
; but ended up thinking this was more informative, in that the surface
; is pretty flat here
;
(add-segments fig2 line10 "black")
(add-text fig2 1 -0.8 "10")
(add-text fig2 0 0.66 "10")
(add-segments fig2 line3 "black")
(add-text fig2 0.5 -0.6 "3")
(add-segments fig2 line1 "black")
(add-text fig2 0.05 0.13 "1")
(add-points fig2 '(1) '(1) "black" 5)
(add-text fig2 1 1 "Optimum" 20)
(add-color-legend fig2 '(-1.9 1) "vir" 0 3600)
I recently found the fortplot library, which seems reasonably well featured. I could use this as a drop-in backend for the Scheme graphics primitives. The Lisp-Stat project, which is a restart of XLispStat and common-lisp-stat, is using Vega-Lite in JavaScript for plotting.
Most of my more interesting statistical stuff is still accessible only from the Sib-pair REPL eg MCMC generalized linear mixed models (GLMMs) and finite polygenic models, extrapolation for MC P-values, or distance correlations. It is amusing to write hooks for these from the Scheme side, but better to generalize them more eg the GLMMs currently don't allow crossed random effects. The other question I am thinking about is the benefits of added precision from rational numbers when dealing with collinear or badly behaved linear models. Though the alternative here is just to add support for multiple precision floats into Sib-pair Scheme. I have used David Bailey's fortran arbitrary precision library in the past, quad precision might be enough anyway.
Edited 20260805 to add call to add-color-legend.
Sib-pair internally uses a Fortran variable metric function minimizer originally published in Applied Statistics by John Koval with downloadable code at several sites . Originally, I wrote a Sib-pair Scheme procedure (maximize!) that calls this directly, but this was slightly awkward since the varmet Fortran subroutine has to call back to Scheme to evaluate the objective function repeatedly. I simplified this by requiring this to be in the top level environment.
The other more flexible way is to call an optimizer implemented in Scheme. There don't seem to be a lot of these about, so I ended up porting AS319 to Scheme. It again uses the Sib-pair Scheme matrix procedures, that do run happily in other Schemes.
;
; Algorithm AS 319 Unconstrained variable metric function minimization without derivatives.
; Translated from the Fortran code by DLD Jul 2026
; Koval JJ. Variable Metric Function Minimization
; J R Statistl Soc. Series C (Applied Statistics)
; Vol. 46, No. 4 (1997), pp. 515-521
;
(define (grad fun b f0 er)
"(grad fun b f0 g sa er) Estimate gradient of function"
(define (perturb b i val)
"(perturb b i val) Peturb ith value"
(append (list-select b (seq 0 (- i 1)))
(list val) (list-tail b (+ i 1))))
(let* ((npar (length b)) (jcmax (- npar 2)) (ser (sqrt er)))
(let loop ((i 0) (jc 0) (g '()))
(cond
((= i npar) (reverse g))
(else
(let* ((this (list-ref b i))
(h (* (+ (abs this) ser) ser))
(f1 (fun (perturb b i (+ this h)))))
(loop (+ 1 i) jc (cons (/ (- f1 f0) h) g))))))))
(define (varmet fun b gradtl toler maxfn0 plevel)
"(varmet fun b gradtl toler maxfn ifault plevel) Maximize function"
(define (print-iter label ifn f1 b)
(print label ": " ifn f1 b))
(define (update-d d1 h c)
(let ((n (nrow h)))
(let loop1 ((i 0) (d2 0) (newd '()))
(if (< i n)
(let ((s (dotprod (row h (+ 1 i)) c)))
(loop1 (+ 1 i) (+ d2 (* s (list-ref c i))) (cons s newd)))
(list (+ 1 (/ d2 d1)) (reverse newd))))))
(define (update-h d1 h t c)
"(update-h d1 h t c) Update numerical estimate Hessian"
(let* ((n (nrow h)) (n2 (* n n)) (newd (update-d d1 h c)) (d2 (car newd)) (d (cadr newd)))
(let loop1 ((i 0) (j 0) (k 1) (ti (car t)) (di (car d))
(oldh (flatten h)) (newh '()))
(let* ((tj (list-ref t j)) (dj (list-ref d j))
(hterm (- (car oldh) (/ (- (+ (* di tj) (* ti dj))
(* d2 ti tj)) d1))))
(cond ((= k n2) (matrix (reverse (cons hterm newh)) n n))
((= j (- n 1)) (loop1 (+ 1 i) 0 (+ 1 k) (list-ref t (+ 1 i))
(list-ref d (+ 1 i)) (cdr oldh) (cons hterm newh)))
(else (loop1 i (+ 1 j) (+ 1 k) ti di (cdr oldh) (cons hterm newh))))))))
(define ig 0)
(define ifn 1)
(define ler #f)
(define ifault 0)
(define npar (length b))
(define np (+ npar 1))
(define maxfn (if (= maxfn0 0) 1000 maxfn0))
(define w 0.2)
(define icmax 20)
(define f0 (fun b))
(if (> plevel 0)
(begin
(print ";; npar=" npar "gradtl=" gradtl "toler=" toler "maxfn=" maxfn)
(print-iter ";; Init" ifn f0 b)))
(define initgrad (grad fun b f0 gradtl))
(set! ig (+ 1 ig))
(set! ifn (+ ifn npar))
(if (> plevel 0) (print ";; Initial g=" initgrad))
(if (> ifn maxfn)
(error (paste "More (" ifn ") parameters than maximum allowed ("
maxfn ") function evaluations!")))
;;;; label 10 reset h to identity matrix
(let mainloop ((lik f0) (b b) (g initgrad) (h (diag-matrix npar)) (ilast ig))
;;;; outer iteration
(let dloop ((i 0) (d1 0.0) (b b) (g g) (h h) (t '()))
(let ((d b) (c g))
(if (< i npar)
(let ((s (- (dotprod (row h (+ i 1)) g))))
(dloop (+ i 1) (- d1 (* s (list-ref g i))) b g h (append t (list s))))
(if (≤ d1 0.0)
(if (= ilast ig)
(list fun b (fun b) ifault) ;; return
(mainloop lik b g (diag-matrix npar) ilast))
;;; label 90 inner iteration
(let loop90 ((ck 1) (ic 0) (d d) (t t))
(let* ((b (map + d (map (lambda (x) (* ck x)) t)))
(icount (sum (map (lambda (x) (if x 1 0)) (map = b d)))))
(if (≥ icount npar)
(if (= ilast ig)
(list fun b (fun b) ifault) ;; return
(mainloop lik b g (diag-matrix npar) ig))
(let ((f1 (fun b)))
(set! ifn (+ 1 ifn))
(if (> plevel 0) (print-iter ";; Iter" ifn f1 b))
(cond ((> ifn maxfn) (display "ifault=4 ") (list fun b f1 4)) ;; return
(ler (if (= ic icmax)
(list fun b f1 3) ;; return
(loop90 (* w ck) (+ ic 1) d t)))
((≥ f1 (- f0 (* d1 ck toler)))
(loop90 (* w ck) (+ ic 1) d t))
(else
(set! f0 f1)
(set! g (grad fun b f0 gradtl))
(set! ig (+ 1 ig))
(set! ifn (+ npar ifn))
(if (> ifn maxfn)
(list fun b f0 4) ;; return
(let* ((t (map (lambda (x) (* ck x)) t))
(c (map - g c))
(d1 (apply + (map * t c))))
(if (≤ d1 0)
(mainloop f0 b g (diag-matrix npar) ig)
(dloop 0 0.0 b g (update-h d1 h t c) '()))))))))))))))))
Here it is in Guile, minimizing the Rosenbrock banana function:
(load "stats.scm") (define (rbf x) "Rosenbrock banana function" (+ (square (- 1 (first x))) (* 100 (square (- (second x) (square (first x))))))) ;; Known minimum of rbf 0 occurs at (1, 1) scheme@(guile-user)> (varmet rbf (list 0.5 0.5) (expt 10 -18) (expt 10 -10) 0 1) ;; npar= 2 gradtl= 1/1000000000000000000 toler= 1/10000000000 maxfn= 1000 ;; Init : 1 6.5 0.5 0.5 ;; Initial g= -51.00000411775891 50.00000403701854 ;; Iter : 4 729948087.8600423 51.50000411775891 -49.50000403701854 [...] ;; Iter : 103 8.877976567691643e-14 0.9999997020829935 0.9999994036638515 $6 = (#<procedure rbf (x)> (0.9999997020829936 0.9999994036638515) 8.87797656330688e-14 0)
I tried to run this using rational numbers, but it bogged down with larger and larger denominators to the estimates.
Goodnight [1979] is a nice historical review, explaining how it is a refinement on pivoting in Gauss-Jordan elimination. It can be used for solving systems of linear equations, multivariate regression, ordinary square matrix and generalized matrix inversion. In Scheme, once one has defined the usual matrix operations on lists of lists, one can add,
(define (sweep A k)
"(sweep A k) Apply sweep operator kth pivot to matrix A"
(let ((pivot (matrix-ref A k k)))
(if (zero? pivot)
#f
(let* ((d (/ 1 pivot))
(res (matrix-row-set! A k
(map (lambda (x) (* x d)) (row A k))))
(b (list-element-set! (col res k) k 0)))
(matrix-set! (matrix-col-set! (matsub res (outer-product b (row res k))) k
(map (lambda (x) (* x (- d))) b)) k k d)))))
Then, inverting a square matrix is carried out by sequentially sweeping each row:
(define (inverse A)
"(inverse A) Inverse of matrix A"
(let ((nv (nrow A)))
(let loop ((i 1) (A A))
(if (> i nv)
A
(loop (+ i 1) (sweep A i))))))
The nice property of this is that it works well with exact numbers:
scheme@(guile-user)> (inverse '((2 4) (3 1))) $2 = ((-1/10 2/5) (3/10 -1/5)) scheme@(guile-user)> (inverse (inverse '((2 4) (3 1)))) $3 = ((2 4) (3 1)) scheme@(guile-user)> (inverse (inverse '((2.0 4.0) (3.0 1.0)))) $4 = ((2.0 4.0) (3.0000000000000004 1.0000000000000009))
The Longley dataset [1967] is a highly collinear set of econometric indicators used to test numerical accuracy of regression software. NIST gives "certified results" to 15 significant digits, by carrying out the calculations to 500 digits of floating point precision.
;
; GNP.deflator GNP Unemployed Armed.Forces Population Year Employed
(define longley (matrix
'(83.0 234.289 235.6 159.0 107.608 1947 60.323
88.5 259.426 232.5 145.6 108.632 1948 61.122
88.2 258.054 368.2 161.6 109.773 1949 60.171
89.5 284.599 335.1 165.0 110.929 1950 61.187
96.2 328.975 209.9 309.9 112.075 1951 63.221
98.1 346.999 193.2 359.4 113.270 1952 63.639
99.0 365.385 187.0 354.7 115.094 1953 64.989
100.0 363.112 357.8 335.0 116.219 1954 63.761
101.2 397.469 290.4 304.8 117.388 1955 66.019
104.6 419.180 282.2 285.7 118.734 1956 67.857
108.4 442.769 293.6 279.8 120.445 1957 68.169
110.8 444.546 468.1 263.7 121.950 1958 66.513
112.6 482.704 381.3 255.2 123.366 1959 68.655
114.2 502.601 393.1 251.4 125.368 1960 69.564
115.7 518.173 480.6 257.2 127.852 1961 69.331
116.9 554.894 400.7 282.7 130.081 1962 70.551) 16 7))
(define (reg XY)
(let* ((nv (+ (ncol XY) 1)) (W (cbind (matrix 1 (nrow XY) 1) XY)))
(let loop ((i 1) (A (matmult (transpose W) W)))
(if (= i nv)
(col A nv)
(loop (+ i 1) (sweep A i))))))
Below I have used Guile's exact numbers (converting to decimal fractions of tenths or thousandths, then inexact->exact to print the results), and then with the data matrix as inexact floats. I compare this result to the those from the R statistics package's default lm procedure. The R result comes from a QR decomposition using 128 bit floats, and will have at most 17 digits of precision.
| Regression coefficient | Guile 3.0.9 rationals | Guile floats | R 4.6.1 lm |
|---|---|---|---|
| (Intercept) | -3482.2586345958184 | -3482.2586380605635 | -3482.2586345958148 |
| GNP deflator | 0.015061872271373296 | 0.015061872670509922 | 0.015061872271372779 |
| GNP | -0.035819179292591014 | -0.03581917950029635 | -0.035819179292591000 |
| Unemployed | -0.02020229803816825 | -0.02020229806623592 | -0.020202298038168240 |
| Armed Forces | -0.010332268671735919 | -0.010332268676397485 | -0.010332268671735891 |
| Population | -0.051104105653580714 | -0.05110410420426831 | -0.051104105653579195 |
| Year | 1.8291514646135518 | 1.8291514663248238 | 1.8291514646135503 |
The "exact" results match the 15 digits of the NIST Certified Regression Statistics. R misses for the Population variable, getting only 12 digits. Sib-pair Scheme floats gets the same results as the Guile floats.
I was currently unable to run lme4 under R. R has been bumped up to 4.6.1 (2026-06-24), and the existing lme4 gets:
Error: package or namespace load failed for 'lme4' in dyn.load(file, DLLpath = DLLpath, ...): unable to load shared object '/home/R/lib/R/library/rlang/libs/rlang.so': /home/R/lib/R/library/rlang/libs/rlang.so: undefined symbol: SETLENGTHHowever, recompilation of lme4 failed with:
error: 'R_NamespaceRegistry' was not declared in this scope
Dirk Eddelbuettel writes on 2026-April-4:
Now it is time to retire R_NamespaceRegistry for which a pending diff would be to call the new R_getRegisteredNamespace() function.
but elsewhere mentions this causes a few packages to fail installation on Debian.
I fixed this by updating my older versions of Rcpp and rlang, and they have been resolved by the lme4 maintainers by a change to the dependencies.
The Ackerman function is a famous older benchmark for scripting and interpreted languages. On tinonee, it takes:
| Program | (ack 3 9) | (ack 3 12) |
|---|---|---|
| Chez Scheme 9.5.8 | 0.07 s | 2.7 s |
| Gfortran 13.3.0 -O3 | 0.06 s | 3.0 s |
| Flang-20 -O3 | 0.09 s | 4.9 s |
| chibi-scheme | 0.9 s | 59.0 s |
| ribbit compiled to C | 1.8 s | 292.7 s |
| scm 5e3 | 5.0 s | >320, segfault |
| ribbit repl | 44.3 s | - |
| Sib-pair Scheme | 153.4 s | >2900 s |
Added gamma regression to "fpm". The "stratified" command accepts a sampleweight variable. The "out" and "print" commands output now varies column width for each categorical variable's width. The Scheme (dgamma) procedure returns log density, while the (maximize!) procedure allows one to maximize a Scheme function taking a list as its one argument. The log10 procedure calls that Fortran function.
The "get" command can now be on a stratifying "group" variable. Sib-pair Scheme also can (delete) a variable in the top level environment, and added the bitwise operations bitwise-and and friends, arithmetic-shift, bit-count (ie POPCNT). Fixed up inexact->exact for big real numbers, and number->string for other bases. Hash tables made available for Scheme, and can be used from Sib-pair as an associative array eg "%area["United States"]" and set as "macro area["United States"]=9629091.
The (gllm) scheme function carries out log-linear modelling for complete and incomplete cross-classified tables of counts. Can be used, for example, for latent class analysis. This is an EM algorithm using IPF [Haber 1984]. Sib-pair uses EM with a Poisson GLM internally to implement a few analyses eg "hapassoc". The (gllm) allows the analyses my old LOGLIN program and R gllm package did, though not currently bootstrapping. The (mode) gives an NPMLE of the mode fitting a unimodal density model to a vector. This was already used internally for summarizing MCMC models.
Pinillos puts forward an argument that a humanities type education will give good judgement that can be applied to problems of living, and compares it to chess knowledge.
But, chess players also agree on who is a better player, and also on what is an aesthetically interesting move or line. And chess players have also defended devoting say 10000 hours of one's young life on gaining expertise on intrinsic grounds, and on benefits to general practical cognition. I was always struck as a kid by Ian Fleming's cunning chess master (a quick search tells me he is Kronsteen) who runs SPECTRE - why else did ancient aristocrats have their children play a war game?
After kernel update yesterday (Ubuntu 24.04.4 LTS, 6.8.0-124-generic), the nouveau driver was supplanted by a newly installed nvidia-535 driver, which does not support my antique GT730, so my display went to the fallback llvmpipe. The nvidia-470 "legacy" driver (which does support older cards) was listed as present, but when I tried to uninstall the 535 driver, this was blocked because of a dependency of the 470.
Eventually, I found someone with the same problem, who resolved it by purging all the NVIDIA packages:
sudo apt purge 'nvidia*' sudo apt autoremove
I was originally planning to retain the proprietary 470 driver, but all deleted now. Apparently, Ubuntu will be dropping support anyway. I have compared nouveau to nvidia-470 previously on my particular setup - no real differences.
Compiling Sib-pair: gfortran -g
| Core i7 820 2.56 GHz [GeForce GT 730] 4 GiB | 58.9 s |
| Pentium CPU 6405U @ 2.40GHz [UHD Graphics 620] 4 GiB | 54.1 s |
flang-new-20 -O2:
| Core i7 820 2.56 GHz [GeForce GT 730] 4 GiB | 281.7 s |
| Pentium CPU 6405U @ 2.40GHz [UHD Graphics 620] 4 GiB | 271.1 s |
Another guitar-drum math rock duo, from Saguenay, Canada. Looping, microtonal tuning.
Finally added bignums to Sib-pair Scheme. This is (at this time) using slowish algorithms from Chapter 1 of Brent & Zimmerman's Modern Computer Arithmetic. I had put this off because I really don't have that much use for the added precision
I've also slowly been adding in more and more XLispStat type facilities. For example, one can replicate the R cars example for lowess, with plotting using the groovy EGGX graphics library.
(require 'plot) (define ys (lowess speed dist 'frac (/ 2 3))) (define ys2 (lowess speed dist 'frac 0.2)) (define fig1 (create-plot speed dist 500 "speed" "dist")) (add-points fig1 speed dist "red") (add-lines fig1 speed ys "blue" 0) (add-lines fig1 speed ys2 "blue" 1) (plot-title fig1 "Lowess for cars dataset") (add-lines fig1 '(2 5) '(130 130) "blue" 0) (add-lines fig1 '(2 6) '(120 120) "blue" 1) (add-text fig1 6 130 "f=2/3" 14) (add-text fig1 6 120 "f=1/5" 14)
And I have added a scheme interface allowing full general log-linear modelling, including latent class analysis, as I previously had available in another older Fortran program.
Andrew Gallant describes his implementation and tests, and pointed to the earlier post by Mike McCandless who summarizes:
...finite-state machines that map a term (byte sequence) to an arbitrary output...Essentially... a SortedMap<ByteSequence,SomeOutput> They generally support the same operations as FSMs (determinize, minimize, union, intersect, etc.). You can also compose them, where the outputs of one FST are intersected with the inputs of the next, resulting in a new FST.
Gallant shows they are compact ways of storing searchable indexes eg all unique words in gutenberg.org was 41 MB in plaintext, and reduced to a 22 MB FST (gzip compressed 13 MB), and the Archive's DOI list was 113 MB as an FST, and 176 MB gzipped. They were also up to 10-fold slower in retrieving than a hash...
With N and V.
High tides:
Jumpinpin Bar 10.44 1.7 Cabbage Tree Pt 12.44 1.5
Usual route Northern end Tabby Tabby and between Mosquito and Short, then Duck Creek. Back past Eden and Southern end of Tabby Tabby, re-encountering the sand banks between Cabbage Tree Point and Tabby Tabby.
Left ~11.30
Swan Bay 1420-1515
Back 1815
There are lots of similar tools around, but I felt the urge to roll my own, and they have just been added to Sib-pair as utility commands. The anagram command generates anagrams and subanagrams for the given set of letters, defaulting down to subwords of size 4. One can request a phrase of given target lengths by slashes. The words command greps the dictionary so allowing missing letters with lists of included or excluded possibilities, but one can constrain equality between letters by specifying a pattern string eg
>> words */..l /11231/11./ eerie eel
Obviously this is for "Codecracker" type puzzles. I use a few freely available system dictionaries, including one for hyphenated terms. The combinatorics stuff is adapted from the fortran 77 examples in NijenHuis and Wilf's 1978 textbook Combinatorial Algorithms For Computers and Calculators, which is fun because all the recursion is done iteratively, so very helpful for me to try and understand. In passing, I once compared Fortran recursive functions versus the well known Stalin Scheme compiler on some little benchmarks - gfortran was faster.
Need it to protect simple minded printing and evaluation.
The SRFI-1 version is
(define (circular-list? x)
(let lp ((x x) (lag x))
(and (pair? x)
(let ((x (cdr x)))
(and (pair? x)
(let ((x (cdr x))
(lag (cdr lag)))
(or (eq? x lag) (lp x lag))))))))
And the Fortran looks pretty similar:
function iscirc(p)
logical :: iscirc
integer :: p
integer :: lag, x
x=p
lag=p
iscirc=iscirc2(x, lag)
end function iscirc
recursive function iscirc2(x, lag) result (res)
logical :: res
integer :: x, y, z, lag
res=ispair(x)
if (res) then
y=cdr(x)
res=ispair(y)
if (res) then
z=cdr(y)
lag=cdr(lag)
res=(z == lag)
if (.not.res) res=iscirc2(z, lag)
end if
end if
end function iscirc2
Meta are developing a data compression framework that combines different compression algorithms in an optimal way for various types of target data.
For a Sib-pair binary dataset, gzip gets a 90% size reduction
| release9_imp95_mc1r.bin | gzip | 2060963 | 201900 |
$ zli list-profiles Available profiles: -| csv = CSV. Pass optional non-comma separator with --profile-argFor an R data image. -| i8 = Signed 8-bit data -| le-i16 = Little-endian signed 16-bit data -| le-i32 = Little-endian signed 32-bit data -| le-i64 = Little-endian signed 64-bit data -| le-u16 = Little-endian unsigned 16-bit data -| le-u32 = Little-endian unsigned 32-bit data -| le-u64 = Little-endian unsigned 64-bit data -| numeric-ml-selector-64 = 64 bit numeric data using ml selectors (Placeholder) -| parquet = Parquet in the canonical format (no compression, plain encoding) -| pytorch = Pytorch model generated from torch.save(). Training is not supported. -| sao = SAO format from the Silesia corpus -| sddl = Data that can be parsed using the Simple Data Description Language. Pass a path to the data description file with --profile-arg. -| sddl2 = Data that can be parsed using Simple Data Description Language v2. Pass a path to the pre-compiled bytecode file with --profile-arg. -| serial = Serial data (aka raw bytes) -| u8 = Unsigned 8-bit data $ zli compress --profile serial release9_imp95_mc1r.bin -o release9_imp95_mc1r.bin.lzi Compressed 2060963 -> 185582 (11.11x) in 37.212 ms, 55.38 MB/s $ zli train --profile serial release9_imp95_mc1r.bin -o binner.zlc [...] Trained 1 compressors in 1.878941 minutes (wall time). Benchmarking trained compressor... 1 files: 2060963 -> 179045 (11.51), 47.77 MB/s 837.06 MB/s Training improved compression ratio by 3.65%
$ ls -l moles.RData -rw-rw-r-- 1 davidD davidD 556913172 Mar 4 15:55 moles.RData $ time gzip moles.RData real 0m23.771s user 0m21.009s sys 0m1.019s $ ls -l moles.RData* -rw-rw-r-- 1 davidD davidD 532777662 Mar 4 15:55 moles.RData.gz $ time zli compress -profile serial moles.RData -o moles.RData.lzi Compressed 556913172 -> 535308740 (1.04x) in 2363.586 ms, 235.62 MB/s real 0m7.926s user 0m2.868s sys 0m2.654s
FGMP is Mark Henderson's implementation of a subset of the GMP API in ~1000 lines of C. Erich Gallesio mentions it as an alternative for bignums in STk.
The view I want to defend starts with a commitment to disquotationalism. In this framework, "truth" is not a property we discover in the world (like "redness" or "solidity"). Instead, it is a tool we invented to solve a specific grammatical problem: blind ascription. Normally, if I want to agree with you, I just repeat what you said. If you say "Snow is white," I say "Snow is white." [I'd like to say:] "For every sentence x, if you said x, then x", [but this] is ungrammatical because the second "x" is in a position that requires a sentence, but it's being used as a variable. To fix this, we created the "true" operator. It allows us to move from the name of a sentence back to the content of the sentence.
What about:
If I want to agree with you, I just repeat what you said. If you say "I never tell the truth," I say "[Yes, I agree,] you never tell the truth." "Sure, but what I just said was a bit paradoxical". "[Yes, I agree,] what you just said was a bit paradoxical".
Via HN: https://karpathy.github.io/2026/02/12/microgpt/.
A NN exemplifying a LLM in 200 lines of Python running over a toy dataset. Others promptly sped it up 60-fold by porting to other languages eg Rust, C++, Julia.
Transformer=Attention+multilayer perceptron: as built up in layers:
Bigram count table - no neural net and gradients
MLP + manual gradients (numerical & analytic) + SGD
Autograd (Value class) - replaces manual gradients
Position embeddings + single-head attention + rmsnorm + residuals
Multi-head attention + layer loop - full GPT architecture
Adam optimizer - this is train.py
Jeffreys discussed the tramcar problem in the 1939 edition of Theory of Probability [Robert et al 2009 Stat Sci 2:141]:
...a man traveling in a foreign country has to change trains at a junction, and goes into the town, the existence of which he has only just heard. He has no idea of its size. The first thing that he sees is a tramcar numbered [m=]100. What can he infer about the number [n] of tramcars in the town? It may be assumed that they are numbered consecutively from 1 upwards.
The standard noninformative prior π(n) ~ 1/n, so P(n > n0|m) ~= m/m0, and the posterior median is ~2m. The MLE is m, and is always below the true value of n.
The Binomial n problem is if multiple observations (r) are present, but power is still not great when the sampling fraction p is unknown. In keeping with this, the prior distribution for p has big effects on the likelihood.
LR(p,N1,N2) = (1-p)(N1,N2)r Π (i=1,r) [ CN1 yi / CN2 yi]
With known p, the generalized Bayes estimator of N with improper prior π(n) ~ 1/n is m/p [Sadooghi-Alvandi 1986].
Added 20260608: Also known as the tank problem.
prompt="An image of dragonfruit and mangoes on a plate in the style of Cezanne"
The protagonist is Yeomyeong, which is one word for Dawn (another is saebyeok).
A CLI text-to-speech tool using the Kokoro-82M open weight TTD model, a large speech language model.
It offers US and UK English, Japanese, Mandarin Chinese, Spanish, French, Hindi, Italian and Brazilian Portugese voices.
Following suggestions on the Korean Wiki Project:
a : "a" in "father" e : "e" in "bed", or beginning of "ai" in "main" i : "ea" in "mean" o : "o" in "boat" or "eau" u : "oo" in "moon" ae : A diphthong, pronounced like "ay" in "say" (young speakers) or "e" in "met" ya : Similar to the "ya" in "yard" eo : "ou" in "young" but "or" in "north" in Seoul eu : similar to "eo" in many dialects yeo : Similar to the "yo" in "yolk" wa : A diphthong, pronounced like "wa" in "water" wae : A diphthong, pronounced like "way" in "away" oe : A diphthong, pronounced like "we" in "wet" ch : "dge" of "hedgehog" between voiced sounds, or "ch" of "chin" j : "ch" of "each" k : "gh" of "doghouse" between voiced sounds, or "c" of "call" otherwise. t : "dh" between voiced sounds, else "t" of "tall"
"My intuition tells me my intuition is very reliable."
Yvette Young is a math rock guitarist (Rolling Stone "155th greatest guitarist of all time") fond of tapping and alternative tuning. Leads band Covet.
Northern start of Scenic Rim Great Walk. 18 km return, 580m ascent. Pleasant.
Started 09:45
Lunch at Mt Mistake camp site 13:15 (15 min)
Finish 16:15
Slowly climbing track to cutesy minigorge with rock steps up little waterfalls, then traverse up past cliffs to summit ridge behind which refurbished logging road from Laidley Gap. Great views of Beau Brummel and Mt Castle and to Brisbane and Maroon from many places along track. Track mainly in private Nature Refuge, with side track to Spicers Hidden Vale private lookout (stairs up onto crag) and their Spicers Mt Mistake Retreat just behind Sunrise Lookout. Thunder shower at 12:30-13:00 while approaching Sunrise Lookout, but dry by return past there.
Vocal Type: Dramatic Soprano
Vocal Range: 4 octaves G2#-G6# [or elsewhere Eb7 - G#5]
Vocal Pluses: Incredibly versatile, rich and emotive voice that has a dark, weighty timbre. The midrange is solid with a warmth and sweetness. The belting range is clear and robust with a fiery passion and connects seamlessly to the head voice. This part of the range is bright, resonate and has the flexibility to sound operatic, contemporary or even able to mimic birdsong. Expert control over the voice means that notes can be sustained effortlessly- with or without vibrato- and pitched perfectly.
Vocal Negatives: Unique and individual singing style is not to everyone's taste.
On reddit:
For comparison, Björk has E3 - B6, Fiona Apple has B2 - E6 and Joanna Newsom has E3 - E6 What I would say in regards to Kate's voice isn't that her vocal range is extraordinary itself (although it is rather good), it's more that it was so comfortable in its upper register, as her tessitura was very high.
That doughnut shaped bowl under old style cooktop hotplates. Our big one 195mm diameter with 50mm central hole (7.63" "Style D").
Just north of Samford in Brian Burke Reserve Nature Reserve (Burke was a Pine Rivers Councillor and Deputy Mayor). Fire trail along 3km of ridge. Scant views, weedy.
Linda Perhacs, Connie Converse, Vashti Bunyan
https://asthmatickitty.com/artists/linda-perhacs/
Horse Lords are an American avant-garde rock band from Baltimore. The members are Andrew Bernstein (saxophone/percussion), Max Eilbacher (bass/electronics), Owen Gardner (guitar), and Sam Haberman (drums)...The band uses the just intonation tuning system.
Fn+W gaming WSAD+arrows Fn+Win blocks start menu etc (Fn+F1...F12 MS commands) Fn+ScLk rainbow keyboard backlight Fn++/- rainbow flash rate Fn+Esc blue backlight without flashing Fn+PgUp brighter Fn+PgDn dimmer Fn+APP side light patterns Fn+Home backlighting pattern (repeat to off) Fn+INS/HOME/PGUP/DEFL/PGDN Switch lighting effect Fn+END programmable (?)
Up to Low Head on the Tamar (lighthouse, pilot station, telegraph cable landing). Then back to Hillwood (near the Batman Bridge) to climb at The Chessboard. Made a meal of Like Nectar for Butterflies (12m, 14), where someone asked if I was aid climbing, and Climbing Date (8m, 12). Cracks between big plates were narrow and shallow.
scrcpy controls an Android phone via USB cable and adb. Go into Developer mode by tapping <Build> the magic 7 times, then turning on USB debugging in the newly visible Developer menu.
Also installed hoardy-adb for reading and manipulating adb backups.
Note that many apps, notably Osmand in this case, do not allow backup by default. Osmand does allow export of maps etc as .osf file.
Stable Diffusion is a Latent Diffusion model developed by researchers from the Machine Vision and Learning group at LMU Munich, a.k.a CompVis.
The most popular image-to-image models are Stable Diffusion v1.5, Stable Diffusion XL (SDXL), and Kandinsky 2.2. The results from the Stable Diffusion and Kandinsky models vary due to their architecture differences and training process; you can generally expect SDXL to produce higher quality images than Stable Diffusion v1.5
from diffusers import DiffusionPipeline
import torch
pipeline = DiffusionPipeline.from_pretrained("stable-diffusion-v1-5/stable-diffusion-v1-5", safety_checker = None)
pipeline = pipeline.to("cpu")
from diffusers import DPMSolverMultistepScheduler
pipeline.scheduler = DPMSolverMultistepScheduler.from_config(pipeline.scheduler.config)
prompt="An image of a squirrel in Picasso style"
image=pipeline(prompt,num_inference_steps=20).images[0]
image
image=pipeline(prompt,num_inference_steps=20).images[0]
image
After 25 minutes per image:
| 
|
Default is usually 50 denoising iterations, but 20 is apparently sufficient for the DPM Solver Multistep Scheduler.
50/50 [1:11:49<00:00, 86.19s/it] [W NNPACK.cpp:64] Could not initialize NNPACK! Reason: Unsupported hardware.
This is because my older x86 cpu does not support AVX2 instructions, with one solution (for later versions of torch) being
import torch torch.backends.nnpack.enabled = False
prompt="a cat" image = pipeline(prompt, num_inference_steps=20).images[0] 5/5 [09:53<00:00, 118.64s/it] Potential NSFW content was detected in one or more images. A black image will be returned instead. Try again with a different prompt and/or seed.
Latter message can be avoided using:
StableDiffusionPipeline.from_pretrained( "./stable-diffusion-v1-5", safety_checker = None)
image = pipeline("a happy black cat", num_inference_steps=5).images[0]
image.show()
image.save("Desktop/blackcat1.jpg")
generator = torch.Generator("cpu").manual_seed(123)
image = pipeline("a happy black cat", generator=generator,
num_inference_steps=5).images[0]
image.save("/home/HTML/Weblog/blackcat2.jpg")
generator = torch.Generator("cpu").manual_seed(431)
image = pipeline("a happy black cat", generator=generator, num_inference_steps=10).images[0]
| 10/10 [15:39<00:00, 93.99s/it]
image.save("/home/HTML/Weblog/blackcat3.jpg")
from diffusers import DPMSolverMultistepScheduler
pipeline.scheduler = DPMSolverMultistepScheduler.from_config(pipeline.scheduler.config)
generator = torch.Generator("cpu").manual_seed(441)
| 10/10 [14:04<00:00, 84.43s/it]
image.save("/home/HTML/Weblog/blackcat4.jpg")
| 
|
| 
|
pipeline.scheduler.compatibles [<class 'diffusers.schedulers.scheduling_euler_discrete.EulerDiscreteScheduler'>, <class 'diffusers.schedulers.scheduling_ddpm.DDPMScheduler'>, <class 'diffusers.schedulers.scheduling_euler_ancestral_discrete.EulerAncestralDiscreteScheduler'>, <class 'diffusers.schedulers.scheduling_dpmsolver_multistep.DPMSolverMultistepScheduler'>, <class 'diffusers.utils.dummy_torch_and_torchsde_objects.DPMSolverSDEScheduler'>, <class 'diffusers.schedulers.scheduling_k_dpm_2_discrete.KDPM2DiscreteScheduler'>, <class 'diffusers.schedulers.scheduling_unipc_multistep.UniPCMultistepScheduler'>, <class 'diffusers.schedulers.scheduling_deis_multistep.DEISMultistepScheduler'>, <class 'diffusers.utils.dummy_torch_and_scipy_objects.LMSDiscreteScheduler'>, <class 'diffusers.schedulers.scheduling_k_dpm_2_ancestral_discrete.KDPM2AncestralDiscreteScheduler'>, <class 'diffusers.schedulers.scheduling_ddim.DDIMScheduler'>, <class 'diffusers.schedulers.scheduling_pndm.PNDMScheduler'>, <class 'diffusers.schedulers.scheduling_heun_discrete.HeunDiscreteScheduler'>, <class 'diffusers.schedulers.scheduling_dpmsolver_singlestep.DPMSolverSinglestepScheduler'>, <class 'diffusers.schedulers.scheduling_edm_euler.EDMEulerScheduler'>]
Note that Tinonee has a NVIDIA GeForce GT730, which has a "Compute Capacity" (or "SM Version") of 3.0. Then torch has to be reverted back to 1.12, but the following does not work:
pip install torch==1.12.0+cu113 torchvision==0.13.0+cu113 torchaudio==0.12.0 --extra-index-url https://download.pytorch.org/whl/cu113
And CPU is Generation 1 (2008) Intel Core i7-920 4 core (TDP 130W)
Running Sharp 32" Aquos had to adjust for overscan using nvidia-settings. Initially this could not save the new xorg.conf. Following instructions from Reddit worked:
sudo chmod u+x /usr/share/screen-resolution-extra/nvidia-polkit
Added 20250922, and on nouveau, something like
xrandr --output HDMI-1 --set underscan on
xrandr --output HDMI-1 --set "underscan hborder" 25 --set "underscan vborder" 25
According to this , the resizing is in the xorg.conf under the "metamodes" option, where one specifies the viewportout to include the margins eg
Option "metamodes" "nvidia-auto-select +0+0 {viewportout=1830x1029+45+25}"or another example where the TV native resolution is 1920x1080,
Option "metamodes" "DFP-0: 1280x720 { ViewPortIn=1280x720, ViewPortOut=1045x675+120+15 }"
A few recommendations. Author, Wo Chi Xi Hong Shi, seems quite prolific. The translater's (Ren Woxing) afterword mentions:
I started this 804 chapter, 3.4 million Chinese character project on May 10, 2014 [and finished Nov 2015], at SPCNET's forums...but I did what I promised, and this resulted in a virtuous (albeit exhausting) cycles of more updates=>more viewers=>more donations=>more updates. On Dec. 22, I formally started Wuxiaworld for Coiling Dragon.
From reddit re S-S:
"Practical Guide to Evil [by ErraticErrata] is my favorite web serial of all time"... Pale Lights is another by EE..."dark and depressing"
Super-Supportive by Sleyca - "launched [Patreon] on May 21, 2023. Within four months they had rocketed to a staggering $25,000 per month earnings. ...[I]s good, really really good, but it is not 8x better [???] than (for example) Thresholder or This Used To Be About Dungeons or Worth the Candle of Alexander Wales."
"Alexander Wales is brilliant, but he's an author's author...and I don't necessarily mean that entirely in a good way. He is very Meta..."
And the r.a.sf.w post:
From: Chris Buckley <alan@sabir.com>
A major reason to write this is just to recommend _Super Supportive_ (review below). Feel free to skip the rest of the message but please read that.
I was reluctant to read many original English webnovels since English authors have so many ways to publish their works. In general the quality ladder goes: webnovels < kindle unlimited < self-published ebooks < publisher ebooks < conventional publishing < award caliber books. Good novels at any rung might be the equivalent of average novels at one or two higher rungs and outstanding ones more, but there's a lot of distance for a webnovel to make up!
The results of my excursion are listed below. Anything listed I would consider a much, much better than average webnovel, but absolute quality will depend on what the read wants. Any other recommendations welcome!
Several of these webnovels I don't hesitate to recommend to this group - not everybody will like them, but I consider them good sf. They are either finished or have multiple novels worth of writing already.
9.5 _Super Supportive_ by Sleyca. https://www.royalroad.com/fiction/63759/super-supportive A modern superhero/supervillain background ala _The Incredibles_ or _Worm_ or _Wild Cards_, but more science fictiony talents than most, set in a galactic civilization with Earth as a newcomer. The MC wants become a supporting superhero, helping other superheros. Overall, I would say that this is the some of the best world-building and character-building I've seen in recent years. So many little things are gotten "right". This is a slice-of-life approach, not action oriented. But one indication of the appeal of _Super Supportive_ to those who like slow, character-driven novels is that this first-time-novelist who started publishing earlier this year has a current Patreon of $25,000/month! Highly recommended (A Favorite)
9 _Mother of Learning_ by Nobody103 A very well done time loop story where the MC is caught in a time loop leading to his death and must figure the causes and consequences both for himself and the world. Completed, and available on Amazon.
8.8 _Valkyrie's Shadow_ by Aeridinae Lunaris Currently a series of 7 novels set in the world of _Overlord_, a Japanese light novel series/anime. The major attraction is the world building and civilization building. A very indepth philosophical look at some of the possibilities in what might happen as very different societies clash in a (mostly) abandoned MMORPG.
The rest of the message is my complete list of reasonable webnovels (additions welcome). Anything above 8 is recommended but all are decent books if you like that genre. Chris
----------------------------------------- Original English Webnovels (includes those that are now Kindle Unlimited) --- Original English Webnovels - Completed
9 Mother of Learning - 108 chapters (3000 pages) - Nobody103 (Domagoj Kurmaic) - Original English(Royal Road), time loop. Nicely done, well thought out world and plot
8.6 Sword God in a World of Magic - 1033 of 1033 - Warmaisach - English but on WebNovel. MC goal is to become powerful enough to kill the God that transmigrated him. Good attention to moral issues as MC is dominated by quest for power.
8 Upon Wings of Change - 60 chapters - CrystalScherer - Wattpad. short, cute novel of reincarnated(?) humans becoming cat-like pets. -- Original English Webnovels - Reading or Writing Ongoing
*9.5 Super Supportive - 109 long chapters (600K words?) - Sleyca - RR - background of superhero/supervillain modern world in SF Galaxy but much more. Among best world-building and character building seen in a long time. Very slow (MC now to be high school (hero-school) sophomore), very well done. Can Sleyca land all the balls in the air that are set up?
*8.8 Valkyrie's Shadow - 7 novels - Aeridinae Lunaris - English RR orig - "Overlord" fan fiction, very in-depth well-thoughtout political philosophy (much more in-depth than "Overlord"). Good characters.
*8.7 Shadow Slave - 1212 - Guiltythree - Webnovel - Isekai, world starts having "Awakened" with aspects that fight/might_become monsters. Well done. Progressive. In Nightmare World starts as slave.
*8.5 Beware of Chicken - v4 ch 58 - CasualFarmer - English orig,RoyalRoad Isekai non-ambitious cultivator gathers strong community.
8.5 A Practical Guide to Evil - Book 2chap43 of 7 - Erraticerrata - Progressive, young girl becomes named Squire (a Villain) to get rid of evil. well thought out
8.5 The Primal Hunter - 729 - Zogarth - RR - Isekai (?), nice progressive system fantasy, rich world/universe.
8.2 The Butcher of Gadobhra - chap 352 - The Walrus King - RR, Game players trapped as contract workers in new virtual reality game, work their way up.
8 A Journey of Red and Black 98 of 224 - Mecanimus - RR, girl becomes vampire English orig, 19th century background
8 Katalepsis - 6.4 of 22+ - Hazel Young - English orig nightmares and hallucinations turn real.
8 Worm - Arc 11e of 30 - Wildbow - Parahuman
7.8 Dungeon Crawler Carl - v3 of 6 - Matt Dinniman - Earth transforms into dungeon for galaxy amusement. Man and cat go deeper, becoming game show stars. amusing, not deep.
-- Original English Webnovels Abandoned
7.5 Metaworld Chronicles - 208 of 491- Wutosama - RR, English orig woman reincarnated as younger version of herself in a parallel world based on magic instead of science.
7.5 Pact - 4..11 - Wildbow - English orig. Need to understand the powers more to enjoy.
7.7 The Perfect Run - ch 31 of - Maxime Durand - Set save points and go back in time. Well done, but couldn't identify with MC.
7.5 Cradle - v 3 of 8 - - MC OK but without agency, world building OK, some plot armor explained, but secondary characters outright poor (only motives are to interact with MC)
7 Portal to Nova Roma - v2 of 3 - - AI isekai, massive plot armor, modern AI setting ludicrous, detailed society development but very unrealistic.
7.8 Super Minion - 52 - Gogglesbear - Hiatus. Lab shapeshifter obtains human.exe and escapes. Nice touches, but not particularly well-written.
7.8 Mage Errant - v3 of ? - John Bierce - Decently done juvenile. Decent magic system but world lacks depth.
8 He who Fights with Monsters - v3 of 10 - Shirtaloon - Isekai. Progressive. MC definitely has "attitude". May appeal to others more than me.
7.7 Taylor Varga - 126 of 307 - - Worm Fanfic. Mary Sue novel. Lots of nice little touches but complete lack of connection to global reality - things that have been publcly displayed would destroy world economy. World wouldn't let their local agreements hold (including the legal contract discussion)
8 Re: Trailer Trash - 51 - FortySixFour - Royal Road - time-travel(regression) slice-of-life. Characterization ranges from excellent to poor (rare). female from 2045 to 1998. realistic. Might come back to it.
8 12 Miles Below - v2 of 3 - Mark Arrows - Mankind mostly exists on surface (exiled) while aware-machines control the depths (down to 12 miles). decently done YA.
Oneida - one album is with Rhys Chatham.
[...In] 2010, when If a Band Plays in the Woods...? was released. This was released with a companion album, If a Lot of Bands Play in the Woods...?, of other bands covering and remixing tracks from If a Band Plays in the Woods...?. Artists on the companion album included The National, Mice Parade, Tokyo Police Club, Tapes 'N Tapes, Jonsi, Mercury Rev, Oneida, We Were Promised Jetpacks, and Frightened Rabbit.
Rainer Werner Fassbinder made a 2-part miniseries of Simulacron-3 in 1973 called Welt am Draht.
The modelsummary package makes pretty, expanded summaries of datasets and output from cor, lme4 etc.
The sucrase-isomaltase or alpha-glucosidase gene (Symbol SI) is 100 kb (48 exons) long, encoding a protein of 1827 amino acids. This precursor protein is cleaved by pancreatic proteases into two enzymatic subunits sucrase and isomaltase that heterodimerize to form the sucrose-isomaltase complex.
Four coding variants are common in European ancestry sucrase-isomaltase combined deficiency, either as a homozygote or compound heterozygote genotype:
| Consequence | Build 37 | rs ID | NFE MAF |
| Val577Gly | 3-164764786-A-C | rs121912615 | 0.0027 |
| Gly1073Asp | 3-164739053-C-T | rs121912616 | 0.0023 |
| Arg1124X | 3-164737443-G-A | rs200451408 | 0.00016 |
| Phe1745Cys | 3-164700803-A-C | rs79717168 | 0.0016 |
| 0.0068 | |||
Consequence is to transcript ENST00000264382.3. NFE=Non-Finnish European.
Inuit of Greenland and Alaska are commonly deficient, with the ancestral frameshift mutation Gly92Leufs*8 seen at a frequency of 17%.
There is a lot of recent interest in whether heterozygotes (that is, individuals carrying one pathogenic allele) experience symptoms, and whether they might represent an undiagnosed proportion of those experiencing irritable bowel or functional dyspepsia symptoms.
The Congenital Sucrase-Isomaltase Deficiency (CSID) Society, a patient/parent group, reports that
[Relatives of] individuals diagnosed with CSID who are heterozygotes have intermediate enzyme values, mild symptoms in infancy, occasional gas, sudden onset of abdominal cramping and pain, and occasional sudden onset diarrhea in childhood. In adulthood most siblings and parents and grandparents exhibit no symptoms but 32% suffer from with symptoms ranging from occasional gas and abdominal cramping, to irritable bowel syndrome and colitis. Most parents report that siblings are not big "sweet eaters" and gas usually follows a day when the child has ingested starch in large amounts at all three meals. Of the 32% who have reported symptoms parents and grandparents have reported complete relief of the symptoms usually associated with irritable bowel syndrome and colitis once they reduce their sucrose intake to less than 2 grams of naturally occurring sucrose per 100 grams of food (without Sucraid) and reduce but not eliminate their overall intake of starch.
The heterozygote (carrier) rate for any of those four important variants is therefore 1.3% in the general European population. There is some evidence that these individuals are more likely to be diagnosed with irritable bowel syndrome [Henström et al 2018; Garcia-Etxebarria et al 2018; Thingholm et al 2019; Husein and Hassan 2020]. Zheng et al [2021], for example, report that UK BioBank participants with a medical record of ICD10-IBS (N=248) were more likely to carry one of 397 rare in silico diagnosed pathogenic SI variants than controls (12.5% v. 8%, OR=1.7)
In Babcock State Park West Virginia, it appears in multiple jigsaw puzzles.
On sci.stat.math, there was a discussion about the reanalysis of a Michelson-Morley type ether drift experiment from the 1930s. I did a quick analysis of the datapoints presented:
#
# https://arxiv.org/vc/physics/papers/0608/0608238v2.pdf
#
png()
nrotations <- 20
nrunlen <- 17
miller <- c(10,11,10,10,9,7,7,8,9,9,7,7,6,6,5,6,7,
7,7,6,5,4,4,4,3,2,3,3,4,1,1,1,0,1,
1,1,0,-1,-2,-3,-2,-2,-2,-1,-1,-2,-3,-3,-5,-4,-4,
-4,-5,-5,-6,-6,-6,-7,-6,-6,-7,-9,-9,-10,-10,-10,-11,-13,
-13,-15,-15,-16,-17,-19,-19,-18,-17,-17,-18,-19,-19,-18,-17,-16,-15,
0,0,0,0,0,0,0,1,4,6,7,8,9,9,10,10,8,
8,7,5,5,3,3,3,4,5,5,5,4,1,0,-1,-1,-2,
-2,-2,-3,-3,-2,-2,-1,-1,-2,-3,-5,-7,-9,-9,-11,-12,-11,
-11,-11,-11,-12,-14,-14,-11, -10, -10,-9,-9,-8,-10,-10,-10,-10,-10,
8,8,8,7,7,6,6,5,4,4,3,1,0,0,-2,-3,-1,
-1,-1,-1,-2,-3,-3,-2,-2,-2,-1,0,-1,-2,-1,0,0,0,
0,1,1,1,1,3,4,6,7,7,7,9,9,9,9,8,9,
9,10,10,10,10,9,9,9,10,10,9,9,9,8,7,7,7,
7,8,9,8,9,9,9,10,11,12,12,12,11,11,11,11,10,
10,10,10,8,5,4,3,3,5,4,3,1,1,0,0,0,0,
0,0,-1,-1,-2,-3,-3,-5,-5,-5,-5,-5,-6,-6,-6,-5,-4,
-4,-5,-5,-4,-5,-6,-6,-5,-5,-6,-6,-6,-7,-7,-8,-9,-10,
-10,-10,-11,-11,-12,-12,-11,-10,-10,-10,-10,-11,-11,-11,-12,-12,-12,
-12,-13,-14,-15,-15,-16,-15,-16,-15,-15,-16,-17,-18,-19,-18,-20,-21,
1,1,2,1,1,2,4,5,7,7,8,7,6,5,4,4,4)
angles <- rep(seq(0,16),20)
repeats <- rep(seq(1,20),each=17)
rezeroed <- c(86, 154, 324)
runs <- rep(1,length(miller))
runs[seq(86,153)] <- 2
runs[seq(154,323)] <- 3
runs[seq(324,340)] <- 4
runs <- factor(runs)
dups <- setdiff(17*seq(1,19)+1, rezeroed)
run1 <- setdiff(seq(1,85),dups)
run2 <- setdiff(seq(86,153),dups)
run3 <- setdiff(seq(154,323),dups)
run4 <- seq(324,340)
nr1 <- length(run1)
nr2 <- length(run2)
nr3 <- length(run3)
nr4 <- length(run4)
nuniq <- nr1+nr2+nr3+nr4
require(locfit)
par(mfcol=c(2,1))
plot(miller[-dups])
fit_pieces <- function(sm=1) {
res <- list()
res$m1 <- locfit(miller[run1] ~ seq(1,nr1), alpha=sm)
res$m2 <- locfit(miller[run2] ~ seq(nr1+1,nr1+nr2), alpha=sm)
res$m3 <- locfit(miller[run3] ~ seq(nr1+nr2+1,nr1+nr2+nr3), alpha=sm)
res$m4 <- locfit(miller[run4] ~ seq(nr1+nr2+nr3+1,nuniq), alpha=sm)
res
}
mod <- fit_pieces(0.4)
lines(mod$m1, lwd=2, col="red")
lines(mod$m2, lwd=2, col="red")
lines(mod$m3, lwd=2, col="red")
lines(mod$m4, lwd=2, col="red")
raw <- miller[-dups]
runs <- runs[-dups]
res <- c(residuals(mod$m1), residuals(mod$m2), residuals(mod$m3), residuals(mod$m4))
dirs <- c(angles[run1], angles[run2], angles[run3], angles[run4])
rotations <- c(repeats[run1], repeats[run2], repeats[run3], repeats[run4])
plot(res, type="l")
abline(v=seq(8,324,8),col="grey80")
par(mfcol=c(1,1))
plot(res ~ dirs, t="p", axes=F, xlab="Markers", ylab="Detrended Residual")
box()
axis(2)
axis(1, at=c(0,4,8,12,16))
axis(3, at=c(0,4,8,12,16),
labels=c(0,expression(pi/2),expression(pi),
expression(3*pi/2),expression(2*pi)))
lines(locfit(res ~ dirs), lwd=5, col="red")
lines(dirs[1:17], cos(4*pi*dirs[1:17]/16), col="blue", lwd=3)
for(i in 1:nrotations) {
lines(dirs[rotations==i], res[rotations==i])
}
require(gamm4)
miller2 <- data.frame(raw,res,runs,dirs,rotations)
miller2$group <- factor(miller2$rotations)
g1 <- gamm4(raw ~ group + s(dirs), random=~(1|group), data=miller2)
summary(g1$mer)
anova(g1$gam)
plot(g1$gam)
m1 <- lmer(raw ~ 1 + (1|group), data=miller2)
m2 <- lmer(raw ~ poly(dirs,1) + (1|group), data=miller2)
m3 <- lmer(raw ~ poly(dirs,4) + (1|group), data=miller2)
anova(m1,m2,m3)
First, the graphical exploration of the data - detrending the data using four localized regressions, but not doing any formal hypothesis testing:
And then showing the detrended data for each rotation of the interferometer (red is localized regression fitted to data, and blue is what Miller was hoping for):
One formal statistical test I calculated was a mixed model with a random intercept for each rotation, and a polynomial fixed effects for the 16 directions within the rotations, which suggests weak support for the idea that the sinusoidal pattern seen in the plots is real.
npar AIC BIC logLik deviance Chisq Df Pr(<Chisq) m1 3 1610.1 1621.4 -802.04 1604.1 Base model m2 4 1585.3 1600.4 -788.66 1577.3 26.7773 1 2.283e-07 *** Linear decline m3 7 1581.7 1608.1 -783.82 1567.7 9.6616 3 0.02167 * Nonlinearity
With usual crowd. From Kalinga Park along Kedron Brook bikeway to Jane St. Up to Patrick Rd, Park Rd, Caesar Rd, Barber Rd, Water Tank Track, then along Cabbage Tree Creek bikeway to Shorncliffe. Back along Sandgate Rd through Nundah tunnel.
Mossio et al [2009] present:
...an organizational account that defines biological functions as causal relations subject to closure in living systems, interpreted as the most typical example of organizationally closed and differentiated self-maintaining systems.
Ellis & Kopel [2019] build on this to argue that "causation of a contextual branching nature" acts in a top-down manner in biological but not other contexts.
There are certain distortions that call for comment because they put in question the applicability of his conclusions to actual translation. First, Quine posits a monolingual native speaker who is evidently incapable of formulating meaning statements in his own language...[T]he radical linguist [is not] given the opportunity to ask him, simply, "What does gavagai mean?" It is clear that for Quine metalinguistic questions and responses inherit the very same limitations as the term they would seek to define. The problem with this from a linguistic perspective is that it is empirically false...Second, Quine stipulates that in order to overcome indeterminacy, the linguist must derive an exact meaning from ostensive reference alone...Thus the linguist is, as it were, at sea without an independent point of reference. Neither of these problems is insuperable in linguistic fieldwork. The linguist gathers a wide variety of evidence for meaning hypotheses, including usage, metalinguistic commentaries, the ways in which the target term combines with others terms in syntactic constructions, analysis of the internal structure of the form (including morphology, compounding, etc.), and grammatical evidence of oppositions between the form and other forms in the language. Similarly, it is not what appears natural in the linguist's native language that guides his or her choice among alternative translations, but rather all that is known of universals of language, and the possible arrays of distinctions that are encoded in lexical forms.
Ironically, the process of successive failed translation may be our best tool in discerning what is specific to any object society or to any "original". In other words, it becomes a method...A second source of equivocation lies in the fact that fully accurate translation is exceedingly difficult, if not impossible, and yet translation is ubiquitous in social life. Bi- or multilingualism, code switching, blending, crossing, paraphrasing, reported speech, and giving accounts are all well-established sociolinguistic phenomena, and all may involve the same key elements as canonical translation...There is a strong line of argument to the effect that understanding is itself a matter of translation: the object understood is translated into some variety of interpretant or representation on the part of the understander...it is itself the basis for understanding.
When Whorf describes how Hopi speakers conceptualize time, or when Boas makes claims about Kwakiutl, they are imagining the native speaker caught in the grips of the native grammar. But what if the Hopi or Kwakiutl speaker is also a fluent English speaker?
IBM version of Julian day, with epoch JD 2299161 (1582-10-15). To convert to ISO, then subtract 141427.
Music-Map is "part of gnod, the global network of discovery", and gives a cloud of musical artists close to a target. For example, Tristan Perich turnd out to be close to Kaitlyn Aurelia Smith, and halfway between Arvo Part and OPN.
My earlier ADE in Wine was fine, but after assorted updates (new SSD - with /home now a link to /main/home, a problem for some programs, and updated Ubuntu) had stopped working. Using winetricks to install ADE 4.5.11 and a rickety dotnet etc lead, for me, merely to "Unhandled exception: 0xe0434352 in 32-bit code (0x7b00dfe2)". This was annoying after buying a book from Kobo, as their Android app would not run on my Nokia 2.2 (Android 11).
A little searching lead to libgourou, which "is a free implementation of Adobe's ADEPT protocol used to add DRM on ePub/PDF files", and can fulfill, download, signin, removeDRM and return loans. It just needed my Adobe ID and password, which I had previously sorted out.
Is Oneohtrix Point Never (OPN). An experimental electronic music composer, performer, singer influenced by "Mahavishnu Orchestra, DJ Premier My Bloody Valentine...Stanislaw Lem and Philip K. Dick."
the no randomness ex nihilo principle, which say together that given an almost-everywhere defined computable map between an effectively compact probability space and an effective Polish space, a real is Martin-Löf random for the pushforward measure if and only if its preimage is random with respect to the measure on the domain.
...we [lack] direct, voluntary control over our motives...lead[ing] to W.D. Ross's objection that there can be "no duty to act from duty"...
[w]e cannot accept want-satisfaction as a final criterion of value because we do not in fact regard our wants as final; ... [rather,] our most difficult problem in valuation is the evaluation of our wants themselves and our most troublesome want is the desire for wants of the "right" kind.
Nancy Lindisfarne and Jonathan Neale:
Female humans have sex year-round. This means...the ratio between sexually active males and females is one to one. In other apes and primates, it varies from 2 to 1 to 40 to 1. That suggests it was easier to create pair bonding and gender equality.The primatologist and anthropologist Christopher Boehm has presented the last piece of the puzzle, in a key article and two influential books. Boehm argues that the equality and sharing among hunter and gatherer bands was culturally and consciously achieved.
[...]
[This is] a coherent picture of a human adaptation to a particular ecological niche evolved over two million years [where] the egalitarian character of this adaptation...is intimately tied to the material conditions of specific environments. Graeber and Wendgrow deal with this impressive range of new material by ignoring it...[exhibiting an] allergy to ecological thinking.
In 2012, the archaeologists Kent Flannery and Joyce Marcus published a brilliant book on The Creation of Inequality. They trace the ways that agriculture has led to inequality in many different parts of the world...
[Informed by] Edmund Leach [1954], Political Systems of Highland Burma, and James C Scott [2009] published The Art of Not Being Governed: An Anarchist History of Upland Southeast Asia. It covered ...the multitudes of rice farmers in the kingdoms of the plains who did run away to the hills. There they reinvented themselves as new ethnic groups of "slash and burn" shifting cultivators. Some of them created smaller class societies, and some lived without class.
A guide to Cornelius Cardew's music [The Guardian]
...he was the victim of an unsolved hit-and-run in 1981 at the age of just 45, and may have been targeted for his leftwing political activism...
The Great Learning is an algorithmic piece (each musician selects using rules as in In C). Revolution Is the Main Trend in the World Today attempts simpler and possibly more accessible music.
Sonic Youth do "Page 183" of Cardew's Treatise on their album SYR4: Goodbye 20th Century. Archive.org. has a live performance of this, as well as several other SY concerts.
With Cathy C. Along A-Break to horse trail (actually grid was safety taped off, presumably given how eroded and steep further down).
> esearch -db snp -query "748487[Base Position Previous] AND 1[Chromosome]" | efetch -format json | jq '{refsnp_id}'
{
"refsnp_id": "184255501"
}
"Previous" is B37 here.
Car shuffle. Left Lawson Rd ~11:15; Lunch on mountain at ~12:10; At car on Parkway 15:20.
Summit ridge very beautiful, with little crags decorated with elkhorns, orchids etc. Hazy views down to Barney etc.
Well-trodden track off lowest point on Lawton Rd as traverses Northbrook Mt to Eagles Nest (462m) and into creek. Followed tape down into gully and Northbrook Ck. Left Creek after ~30 minutes up taped track that ascended old road back to parkway, then 12 minutes down road to usual Northbrook Gorge parking spot (which had 10 cars).
Lang Ya Bang (Wolf's Teeth Club): pole arm 1.7-2.3 m with spiky head. May be used in Wuzuquan Five Ancestors Fist style from Fukien that combines five other styles (including White Crane). Not to be confused with Langya Bang (Nirvana in Fire, a novel and TV wuxia series).
Zhanmadao (Horse chopping sabre [dao]): "a single-bladed anti-cavalry sword...Surviving examples include a sword that might resemble a nagamaki in construction; it had a wrapped handle 37 centimetres long making it easy to grip with two hands. The blade was 114 centimetres long and very straight with a slight curve in the last half."
Xuanxia: Xuan is Daoist immortal, and xi is hero. Genre adjacent to wuxia but less historical/more fantastical. "Protagonists are usually 'cultivators' (xiuzxinhe, xiushì, or xiuxianzhe)". Crouching tigers and hidden dragons are cultivators hiding in the everyday world. Cultivators seek immortality via neidan (inner alchemy), classically by "reducing the multiplicity of phenomena...into the single source (the Dao) from which they arose...the multiple energies of the human body are fused and circulated in cyclical movements (zhoutian, orbits), which pass through two channels [up spine, down anteriorly]...free from obstructions... at the nodal points (guan)...[eg] between the shoulder blades...joining the three energy centres of the body (dantian) where jing (essence), qi (pneuma) and shen (spirit) are progressively refined" [Clark, The Story of Han Xiangzi].
Jianghu (Rivers and Lakes) is the "community of martial artists in wuxia stories and, more recently, outlaw societies like the Triads." [Wiki]
In My Senior Brother is Too Steady [aka Cautious] (Qidian, author is Yan gui zheng chuan ie Get to the Point, trans Atlas Services), the Goddess Nuwa likes:
The Legend of the Condor Heroes, The Three Brothers of the Earth Dragon, The Demon Cult advocates for Organization, Little Li Flying Sword, Chu Liuxiang and Tian Boguang, Drunken Rain in the Field...
The Ming dynasty sees the rise of vernacular novels, including "novels of gods and demons" - shenmo xiaoshuo, many of which are Daoist, and involve the Eight Immortals (baxian): Xiyou ji (Journey to the West), Dongju ji (Journey to the East).
Fēngshén Yǎnyì (The Investiture of the Gods) "is a 16th-century Chinese novel and one of the major vernacular Chinese works in the gods and demons (shenmo) genre written during the Ming dynasty. Consisting of 100 chapters, it was first published in book form between 1567 and 1619...It is a romanticised retelling of the overthrow of King Zhou, the last ruler of the Shang dynasty", and mined by many modern xuanxia authors, including Get to the Point (above).
Jin Yong wrote the 1963-6 newspaper serial Tian Long Ba Bu ie Demi-Gods and Semi-Devils set in the Northern Song (~1000 CE), which follows his The Legend of the Condor Heroes (1957-9).
The Deer and the Cauldron [serialised 1969 to 1972], even among fans of Jin Yong's novels, has divided critical opinion mainly due to the character of Wei Xiaobao...an antihero who relies on wit and cunning. Ni Kuang argued that The Deer and the Cauldron was "the best novel of all time, Chinese or foreign".
Schliesser previously has written on this, but summarizes
I am a fan of a species of generic theories, synthetic philosophy, by which I mean a style of philosophy that brings together insights, knowledge, and arguments from the special sciences with the aim to offer a coherent account of complex systems and connect these to a wider culture, policy, or other philosophical projects (or both). Now the generic theories of synthetic philosophy (I have used as examples game theory, information theory, Darwinianism, actor-network theory, etc. ) are developed within and in conversation with special sciences and need not be topic neutral in a general way. Yet, these generic theories are not just useful in the special sciences, but often can be made action guiding in non-trivial ways.
In passing, there is a big physics literature on heaps (liquefaction or sorting when vibrated, different gravitation etc). If one said, say:
Premise: A heap is collection of objects with an angle of repose.
H1: Removal of one grain of sand in a heap of N grains could cause the heap to collapse ie no member of the collection rests stably and entirely on other grains.
H2: Removal of one grain of sand in a heap of N grains could cause the heap to collapse ie no matter how one rearranged the resulting collection, it would no longer be possible for any member of the collection to rest stably and entirely on other grains.
H3: A minimum heap size might be four.
curl -X 'GET' \
'https://api.ncbi.nlm.nih.gov/variation/v0/refsnp/13014919/frequency' \
-H 'accept: application/json'
But then need the Frequency metadata to work out identity of population: eg SAMN10492703 = "Subjects with non-Hispanic African ancestry"
[The cerebellin 2 gene] CBLN2 is expressed by more types of PFC neuron in humans than in macaques or mice...Shibata et al...discovered...an enhancer...that is active during early PFC development...Comparisons of the enhancer sequence in different species revealed that two deletions that probably occurred between about 7 million and 12 million years ago removed some of the SOX5 binding sites from the genome of the common ancestor of humans and chimpanzees. In cultured cells, the human and chimpanzee CBLN2 enhancers were not suppressed by SOX5, whereas the gorilla and macaque versions of the enhancer were moderately suppressed, and the mouse version, which has more SOX5 binding sites than the primate versions, was most strongly suppressed...mice with the human enhancer had more...dendritic spines...and more synaptic structures in the PFC.
In another paper, these authors highlighted retinoic acid as a key regulator of PFC development, with higher levels in humans, and in the PFC than other regions. Increasing retinoic acid signalling in the mouse "led to the development of more projections between the medial thalamus and the mPFC".
Smits and Monden [2011] report on twinning rates for 76 countries using the Demographic and Health Surveys (DHS).
The average of the national twinning rates in the 76 countries was 13.1 per 1000 (each country weighted equally) or one twin birth in 76.3 births. This figure is close to the average rate of spontaneous twinning mentioned in the literature of one twin birth in 80 births.... A zone with high twinning rates of above 18 per 1000 runs from Guinea in the West along the Atlantic coast to Congo DR and then crosses the continent to Tanzania, Mozambique and the Comoros. South of this zone, in Namibia, Lesotho, South Africa, and Madagascar, twinning rates are clearly lower with values of 11-15... Benin [has] 27.9 per 1000... twinning rates in Latin America turn out to be at a similar low level as in Asia.
Olusanya [2011] reports a Yoruba single-hospital series of confinements where 157/4573 (3.4%) were multiples.
Originally (before 2007) Melbourne band, pronounced "Hate Rock" - now duo. Via Guardian article.
With K. Left 0845. Junction at 0930. Greene's Falls 1110. Coffee at Elm Haus, then ~12:30 along Westside track to Lawton Rd, and down Alex Road. Some surprisingly (in the sense I didn't remember them) tall waterfalls above junction of side creek and Love Ck. Back at car 1620.
The notes for his Sydney Festival contribution:
In the mid 1970s Rhys Chatham was working alongside sonic experimentalists including Philip Glass, LaMonte Young and Steve Reich in New York, when his world was turned upside down by the Ramones' self-titled debut album.
Was involved in No Wave, playing with Glen Branca (recall), and including the Band of Susans members and Nina Canal (punk/noise band Ut) in his ensembles [Wikipedia]. The piece A Crimson Grail is for a 400 guitar orchestra. There are several effects in the latter that remind me of King Crimson (specifically in Starless and The Night Watch).
Commelin and Scholze decided to call their Lean project the Liquid Tensor Experiment, in an homage to progressive-rock band Liquid Tension Experiment, of which both mathematicians are fans.
Includes Jordan Rudess, Tony Levin, Mike Portnoy, John Petrucci. Portnoy and Petrucci were with Dream Theater, which "along with Queensrÿche and Fates Warning...has been referred to as one of the "big three" of the progressive metal genre." LTE formed 1999. Petrucci is "#17 greatest guitarist in Guitar World Magazine".
A response to Bright on the "end of analytic philosophy" comments:
Carnap was clear that his notion of intension...as functions from state descriptions to extensions... wasn't meant to do the work that older notions of intension (sense, comprehension, connotation, content, etc.) were able to do..."intensional semantics" is now, in many idiolects, coextensional with "possible-worlds semantics"...[but] Kit Fine's work on essences in the 1990s and early 2000s, for instance, is predicated on the observation that possible worlds cannot by themselves distinguish the asymmetric ontological dependence (so the thought runs) of the singleton set {Socrates} on the human being Socrates. Each exist at exactly the same worlds, and so the ontological dependence of a set on its members cannot be modelled in those terms. Consequently, models restricted to possible worlds cannot explain the different truth-conditional meanings of "Socrates is essential to {Socrates}" and "{Socrates} is essential to Socrates".Rather than try to build a notion of intension inside model theory, then, it seems more promising to treat model theory and proof theory as two formalisms for reconstructing the old extension/intension distinction concerning complementary notions of meaning - one ontological, and concerned with word-world (and, as I'll note in a moment, world-word) relations, and the other deontic, concerned with word-word relations.
D and K. Left car at park off Woodline Drive ~1300, back at 1700.
Anticlockwise circuit - Yuddamun track, Magical Mystery Tour, off-track then found new track along ridge to Balancing Rock (1500). Nice track right on top of ridge to Frogmouth trail, then Spring Mt trail and Spring Mt summit (1530). Back via Spring Mt trail, Ring Rd, David Savage Trail. Saw only 2 walkers and 2 cyclists.
CAC [coronary artery calcification] is typically quantified using the Agatston score - a sum of the attenuation (in Hounsfield units) and area of all CAC lesions in the coronary arteries - which is then categorized into very low risk (CAC = 0), mildly increased risk (CAC = 1-99), moderately increased risk (CAC = 100-299), and moderate to severely increased risk (CAC ≥ 300).
Akintoye et al [2021] pooled the ARIC, CHS and MESA studies. Of ~14000 participents with low or borderline risk (mean age 55), 38% had CAC (>0 Agatston units).
One very useful tool is the MESA Risk Score Calculator:
According to the Cardiac Society ANZ, "The evidence for pharmacotherapy is less robust in patients at intermediate levels of CAC 100-400, with modest benefit for aspirin use; though statins may be reasonable if they are above 75th centile."
There is plenty of good mathematics in homotopy type theory, which can of course be applied to natural science in the usual way. But when the theory is dressed up as a new foundation for mathematics, it is combined with philosophical claims about the mathematics on which those applications don't really make sense. For example, mathematics is claimed to be interpreted in terms of proofs, but mathematical equations about the physical world can't be proved in the relevant sense. Isomorphic systems are claimed to be identical, but physical systems can have the same structure without being identical
From The Birth of Loop:
Riley frequently played solo all night concerts 1968-9 using keyboards, saxophone and tape delay...Pandit Pran Nath's other students were Don Cherry and Jon Hassell. His Music for The Gift (1963) is the first piece using a delay loop.
From Wikipedia
Acoustic guitar duo Rodrigo y Gabriela covered "Echoes" on their 2019 album Mettavolution, one of 7 tracks which won the album an award for Best Contemporary Instrumental Album at the 2020 Grammy Awards.
D and K. Left Gabba 0900, Glen Rock 1050, Mt Philps 1300 (phone coverage just at summit), back at car 1510. Knee-high grass everywhere with rain (hiding rocks). Followed pad ~20m S along road past information hut that joined correct fire trail (took lower fork after 5 min) and climbed to ridge top at 1200. To summit steep at top, dislodged a few rocks up to 30cm diameter. Dropped down northern ridge to fenceline and track.
Stroll with MG over to base of Graveyard Ridge and up. Where ascend past top cliffs traversed to Tiger Wall Annex. Started 10:40, Caldera at 1500, top of tourist track 1600, car at 17:40.
The Wikipedia article on Expressivism says the Frege-Geach Argument [Geach 1960] is that complex statements involving (embedding) simpler, say, ethical statements eg Modus Ponens "If it is wrong to tell lies, then it is wrong to get your little brother to tell lies" seem to work even though a noncognitivist should find it hard to defend this (Uchii in his example, "if you support the rule of law, then you should not drive without a licence", seems to think this is not a problem). Brandom characterizes the declaration (in naive expressivism) to have imperative force, and testable by substitution into similar complex sentences. He says second-wave expressivism (Blackburne, Gibbard) attempts an answer.
Schroeder [2012] in the Routledge Encyclopedia of Philosophy says:
...Simon Blackburn in Spreading the Word (1984). According to a slight adaptation of Blackburn's central idea, what it is to think that if P, then Q, is just to "boo" thinking that P but not thinking that Q... Though Blackburn's account may be able to explain why some arguments are valid, it has been argued to predict too many arguments to be valid.
Yeah, but consider theological arguments, or astrological inferences - how are they any different?
Zheng et al [2020] mention AUC = qnorm(sqrt(0.5*h2).
And see https://github.com/yandorazhang/CancerEffectSize.
Woodwards's definition of causation underpins his definition of explanation:
An explanation ought to be such that it enables us to see what sort of difference it would have made for the explanandum if the factors cited in the explanans had been different in various possible ways.
He [Woodward 2018 ] and Reutzinger and colleagues [2020] suggest that the factors involved might not have to be causes, in order for an explanation to successfully follow this form, giving a counterfactual explanatory monism.
Older examples of a unitary theory of explanation are Hempel's "covering law" and Kitcher's "unification". Ross [2020] comments that non-causal explanations (under a dualistic model) often involve topological features of the system (eg Konigsberg bridges), but that in fact topology is causal in many scientific settings, even though we normally model these as timeless. Lange [2014,2016] considers that the counterfactuals in a mathematical explanation (ie proofs) are necessarily trivial (quodlibet) in the standard semantics. Baron et al [ 2020] investigate nonstandard truth-closeness-based approaches to the latter, in their example by creating a structural equation model including a counterfactual variable. Since their example was a Euclidean geometry problem, I wonder if the free variable could have been explicitly indexing the geometry or topology in which the problem was set.
Telomere length for all 500,000 participants...
A list of best papers 2010-20
Intel has released its oneAPI toolkit, which includes full support for Fortran 2018. Even better news, it's free if you don't mind "community support". (If you want Priority Support, you can buy it.)
There are two Fortran compilers here. ifort is the next version of what had been in Intel Parallel Studio XE, and is the one with the full Fortran 2018 support. ifx is a new, LLVM-based compiler and is still considered "beta".
https://software.intel.com/content/www/us/en/develop/tools/oneapi/all-toolkits.html
As is usually the case, there may be divots in the F2018 support - these should be reported to Intel. I have been trying the beta and it's very good.
safecopy - rescue data from a source that causes IO errors --stage1 Preset to rescue most of the data fast, using no retries and avoiding bad areas.
Nalls et al [2019] is a meta-analysis of 17 GWAS including UKBB relatives of cases, which "identified 90 independent genome-wide significant risk signals across 78 genomic regions...explain[ing] 16-36% of the heritable risk".
Jacobs et al [2020] use PRS and clinical data in 1267 UKBB cases to assess prediction. Predictors included "a positive family history of PD, a positive family history of dementia, non-smoking, low alcohol consumption, depression, daytime somnolence, epilepsy and earlier menarche".
Pierce:
if one can define accurately all the conceivable experimental phenomena which the affirmation or denial of a concept could imply, one will have therein a complete definition of the concept, and there is absolutely nothing more in itAnd (also via David Matthews (U Adel):
...scientific inquiry is characterised by "contrite fallibilism" and...the scientist is distinguished by "his readiness to dump the whole cartload of his beliefs the moment experience is against them"...truth is "that concordance of a statement with the ideal limit towards which endless investigation would tend to bring scientific beliefs".
Place [1988] suggests if there is "no causal connection between a mental event and the description that is purportedly given of that mental event in the subject's introspective report...it cannot be a genuine description of the event in question". A riposte might be that the verbal description is emitted from the "causal layer". Another might be that introspection is worthless.
To the Davidsonian analysis of event structure, Peter Hacker proposed the example ofHe wisely apologised.Evidently there is a context for such a statement that contains the identity of the apologiser, the act for which the apologising is required, the people who were offended by the act, some idea of how they might seek redress, and so on. A story of a certain shape needs to be supplied. Ranta's ideas on narratives as DTT contexts seem to me appropriate.
How about: "An apology was in order".
Henderson [1982, 2002] Functional Geometry (re-)reconstructs the Escher woodcut in a functional picture language. His R version is at https://mhenderson.github.io/funcgeo/.
A proposition about a imprecisely defined entity can be true about the many almost identical entities you might have specified using more information. Van Fraasen [1966] and Fine [1975] suggest there are supertrue sentences in many cases, while if they are only mostly true there is a "supervaluational truth-value gap". What is the referent for "Mont Blanc"? Are rabbits in holes of that region included?
Among the many groundbreaking developments in the philosophy of language and thought in the last decades motivated by the idea of direct reference, an important one was the clear disconnection of the intentional and the cognitive (and the perceptive). A language user can denote something or successfully point at something, with little or no knowledge of what is denoted. An utterance can refer even while no one knows how to correctly describe its referent. Intentionality is detached from cognition.
Michaelson's SEP summary of reference includes overlapping models: descriptivism ("words refer in virtue of being associated with a specific descriptive content"); causal model; character model (conforming to the rules of reference - "indexical tokens acquire their reference because they are uttered in a particular context, presumably intentionally. But they acquire their reference independently of anything having to do with the speaker's mental state"); intentionalist (a la Grice, so you interact to clarify what we are currently talking about). And Quine, of course.
Assorted quotes
virtue is "a state of character concerned with choice, lying in a mean, i.e., the mean relative to us, this being determined by a rational principle, and by that principle by which the man of practical wisdom would determine it. That we know this "golden mean" by examining the choices made by the man of practical wisdom and also recognize the man of practical wisdom by the fact that he typically selects the mean is a tautology often noted by students of the Ethics.
Thus, we must return to the principle by which the man of practical wisdom operates. Though difficult to operationalize, it is clear. It requires in every particular situation that a balance of what is desirable and what is reasonable be determined through deliberation. "Choice is deliberate desire, therefore both the reasoning must be true and the desire right, if the choice is to be good and the latter must pursue just what the former asserts" (NE 1139a23-26). Although phronesis has universal applicability, it is not "concerned with universals only - it must also recognize the particulars; for it is practical, and practice is concerned with particulars" (NE 1141b14-16). As Ronald Milo observes, Aristotle's notion of good deliberation "presupposes both correct reasoning and reasoning with a view to a good end."
phronesis and moral virtue, or goodness of character, are closely and reciprocally related, in the sense that neither can exist without the other
Aristotle: "perhaps one's own good cannot exist without household management nor without a form of government."
The Greek concept phronesis is famously thematized by Aristotle in the Nicomachean Ethics, and is usually translated as "practical wisdom", as contrasted to technical skill in arts and crafts (techne), the knowledge of science (episteme), the theoretical wisdom of philosophy (sophia), and intuitive reason (nous).
The rulers of the polis are required to render judgment in complicated particular cases in which general laws provide inadequate guidance concerning the right way to proceed, and in which the aim of the decision is the good action and the good life in itself and not some other thing, such as health.
https://pan.ukbb.broadinstitute.org/downloads
Summary statistics MatrixTable: gs://ukb-diverse-pops-public/sumstats_release/results_full.mt (12.78 T)
Meta-analysis MatrixTable: gs://ukb-diverse-pops-public/sumstats_release/meta_analysis.mt (12.54 T)
Essential Tremor (code 1525) N.Cases = 133
From Time magazine.
Slants to younger audiences and best selling series...
Susan Cooper The Dark is Rising
Natalie Babbitt (1975) Tuck Everlasting
Brian Jacques (1986) Redwall (Series, younger)
Diana Gabaldon (1986) Outlander (Series, younger)
Cassandra Clare City of Glass (Mortal Instruments v 3)
Ken Liu
Isabel Ibañez (2020) Woven in Moonlight
James Nicoll suggests Melanie Card (2011) Ward Against Death (series).
With K. Left Yellowpinch 0745. Summit ~12:15. Back at car ~16:10. Two detours from usual route, last being ~100 m from top where traversed W on pad, then technical scramble up below big gully, and had to backtrack slightly to get back onto ridge. Returned Peasant's, so I could collect and drink another 4 litres of water.
A syllogism with an unspoken premise or conclusion. Uncharitably, hiding an unreasonable assumption; charitably, enjoining the listener to perform and agree with the judgment entailed in the hidden material: "rhetoric finds its end in judgment" - Aristotle.
Tang era official 712-770 and the "greatest" of poets (the poet sage, shi sheng), caught up by the An Lushan Rebellion. When Commissioner of Education in Huazhou, a poem includes:
I am about to scream madly in the office
Especially when they bring more papers to pile higher on my desk.
Discussed with interesting commentary at Language Log:
Spring View
Country damaged mountains rivers here/continue
City spring grass trees deep
Feel moment flower splash tears
Regret parting bird startle heart
Beacon fires join three months
Family letters worth ten thousand metal
White head scratch become thin
Virtually about to not bear hairpin
David Hinton's translation:
The country in ruins, rivers and mountains
continue. The city grows lush with spring.
Blossoms scatter tears for us, and all these
separations in a bird's cry startle the heart.
Beacon-fires three months ablaze: by now
a mere letter's worth ten thousand in gold,
and worry's thinned my hair to such white
confusion I can't even keep this hairpin in.
Conal Boyce comments the morphosyllabic language means each line is "like a slideshow". The literal gloss does this, but "lands...somewhere between the New Age Chinoiserie of the 1970s and just plain goofy".
Papers and books of Richard John Lynn .
Via DN, the linguist Chris Potts, discussing GPT-3, contrasts the Lewisian view that "Semantics with no treatment of truth conditions is not semantics" with that of Jerrold Katz, who wrote that "'[t]he arbitrariness of the distinction between form and matter reveals itself', ...argu[ing] throughout his career for an expansive semantic theory in which words and phrases are essentially their own meanings" (generative semantics).
Lewis [1970]:
A meaning for a sentence is something that determines the conditions under which the sentence is true or false. It determines the truth-value of the sentence in various possible states of affairs, at various times, at various places, for various speakers, and so on...[T]he truth-value of a sentence [is] the extension of that sentence; the thing named by a name [is] the extension of that name; the set of things to which a common noun applies [is] the extension of that common noun. [...An] input package of relevant factors [is] an index; any function from indices to appropriate extensions for a sentence, name, or common noun [is] an intension...The plan to construe intensions as extension determining functions originated with Carnap (1947).
Deery "briefly sketches" Putnam-Kripke causal-historical externalism as implying (but perhaps not entailing) essentialism, in that reference to a natural kind is to a natural kind ("essential nature becomes an empirical matter").
Baud and Durand [2012]
(1) financialization of objectives refers to the implementation of shareholder value norms, whose concrete consequences are an increase of the financial flows from non-financial corporations to the financial sector; (2) financialization of investment refers to the increasing share of financial assets owned by non-financial firms; and (3) financialization of operations refers to the development of financial activities and relationships, notably offered to customers and/or imposed on workers and suppliers, by non-financial firms.
"From the beginning of the 1990s up until 2007, leading retailers experienced a slowdown of their growth in domestic markets and yet generated an opposite upward trend in return on equity (ROE)...[via] foreign expansion, financialization of assets, deterioration of suppliers' and workers' positions and the use of working capital management to transform market power into financial gains."
https://www.nature.com/articles/s41467-020-18085-5
The CORE GREML approach fits the Cholesky decomposition of kernel matrices in an LMM, to estimate the covariance between a given pair of random effects. Using extensive simulations, we show that CORE GREML outperforms GREML, providing estimates of variance and covariance components that are free from bias due to correlated random effects.
They report results from a GTE model that "decomposes phenotypic effects into the random effects of the genome and the imputed transcriptome and residuals, i.e., y=g+t+e, [where] the imputed PrediXcan transcriptome consists of expression levels of 227,664 genes from 43 non-sex-specific tissues" [Gamazon et al 2015]. The GTE model always fitted better than a GE model, and better than a GGTE model, where GT was derived from the unweighted genotypes of the 1316391 SNPs used to impute the transcriptome values. Unsurprisingly, G and T are correlated.
In passing, Shor et al [2019] compared their Haseman-Elston based program to Wombat, finding the latter 13 times faster for estimating heritability (AE model) for a dataset containing 250000 records, but crashing when they tried 500000 records. However their package could handle epistatic components without matrix inversion.
Campbell Brown in Consequentialize This:
To 'consequentialize' is to take a putatively nonconsequentialist moral theory and show that it is actually just another form of consequentialism... My strategy is to decompose consequentialism into three conditions, which I call 'agent neutrality', 'no moral dilemmas', and 'dominance', and then to exhibit some moral theories which violate each of these....You say you're concerned about the guy's rights? No worries; we'll just build that into your theory of the good. Then you can be a consequentialist too. And just like that, you've been assimilated.
Dreier is one such, because it is commonsensical. And Jennie Louise [2004] says "[s]ince we are now all under the consequentialist umbrella, the question now becomes not whether we should be consequentialists or not, but whether we should be value-neutralists or value-relativists". Schroeder [2017] says the "Compelling Idea" behind consequentialism is that "it is always permissible to bring about the best outcome...This attractive structure, though, is paired with deeply counter-intuitive verdicts on individual cases". Douglas Portmore states as a maxim that we have more reason to perform one action over another iff, and in virtue of the fact that, we have more reason to desire the outcome associated with the former (disputed by several authors). One problem with consequentializing is that the transformation may be ugly eg absolute injunctions imply infinite value. Colyvan et al [2010] specifically ask whether decision theory can successfully represent the motivation and reasoning of "agents who subscribe to [nonconsequentalist] ethical theories", and mentions Hajek's [2003] hyperreal models of Pascal's Wager (eg avoiding where any mixed strategies still have infinite support).
Yet another interesting interview by Richard Marshall, where Darren Bradley talks about Carnap, formal epistemology, Bayesianism, many-worlds.
Let's start with relevant alternatives theories of knowledge. The idea is that you don't simply know that p; you know that p rather than q. For example, you don't know that grass is green; you know that grass is green rather than white, black, red etc. This theory is useful for answering skepticism about the external world. The skeptic points out that you might be a brain-in-a-vat, and infers that you don't know you have hands. The relevant alternatives theorist can acknowledge a sense in which the skeptic is right - you don't know you have hands rather than being a brain-in-a-vat. However, you know you have hands rather than having hooks.Yes I think that a debate in recent philosophy of biology was solved by Carnap. The point can be made using the slogan that natural selection operates through the survival of the fittest. Some object that "survival of the fittest" is an a priori tautology, so can't be a law of nature. My position is that it is an a priori tautology and also a law of nature...
For a simpler example, think about carburetors. What it is to be a carburetor is to be a device for mixing air and fuel. Consider "the carburetor mixes the air and fuel". That's an a priori tautology, as it follows from the definition of "carburetor". But it can still be a law, and it can still be explanatory. If I don't know that there is a device that mixes air and fuel, or don't know that air is needed for the engine to work, it could be explanatory to be told "the carburetor mixes the air and fuel".
In Bradley [2018], he argues why philosophers should prefer simpler theories against Huemer [2009], who had put forward four counterarguments, where what might work in the sciences will be inadequate for metaphysics etc: Empiricist: science lacks criteria to show that theories preferred as simple are really better; Likelihood: likelihood-based tests fail for philosophical hypotheses eg dualism; Numerousness: larger number of complex hypotheses; Bounded Asymmetry: alternative theory complexity can only go up [?].
Huemer would have been better off just claiming Knightian uncertainty, it seems to me. In another 2018 publication, he argues that Carnap's dismissal of metaphysics is because Carnap felt metaphysical hypotheses could not be justified, rather than verified. Bradley's asks what conceivable evidence could support S over ~S, commenting that in "many cases, metaphysicians engaged in a debate intend that their competing hypotheses should be empirically equivalent":
1C. If we can have justification for S over T then evidence which would support S over T is conceivable.
2C. It is not the case that evidence which would justify S over T is conceivable.
3C. Therefore we cannot have [epistemic] justification for S over T. [though pragmatic justification is quite possible]
This reminds me of Hume on miracles.
From a review of Hardt's Economics Without Laws, where Hands writes:
The first major theme is that not only have philosophers moved away from the Humean, event-regularity, view of scientific laws that dominated the literature of the nineteenth century, they have also steadily moved away from the syntactic view of scientific theories associated with logical empiricism: theories as axiomatized systems of sentences linking theoretical terms and observational terms, generally through some version of correspondence rules. There are many aspects to the recent philosophical developments Hardt discusses, but three of the most important are (i) an increased concern with the actual practice of science (I would call this the naturalistic turn), (ii) an increased emphasis on scientific models and a decreased emphasis on scientific theories, and (iii) the development of various versions of the semantic (or model-based, or non-statement, or predicate) view of scientific theories in which scientific theories are simply families of models.
And,
Under the believable world conception "a model is an entity containing mechanisms that are believed to be similar to the ones operating in the real world" and they "are similar because a model of a mechanism demonstrates the reality of a mechanism by isolating it" (148). Hardt's believable world conception draws heavily (with proper recognition) on both Cartwright's nomological machines and Robert Sugden's notion of a credible world or model (Sugden, 2000; 2009; 2011), but is not identical to either. Hardt summarizes his account of the relationship between idealized economic models and scientific explanation toward the end of the book:models explain by depicting structures which enable the workings of mechanisms, or models are just mechanisms' descriptions. And such models of mechanisms produce beliefs about the real world and thus those beliefs are always true in models producing them. Here my answer to the question of how models explain refers closely to Cartwright's idea of models as blueprints for nomological machines that produce "insights" about regularities in the actual world. (150-151)
And points to Marc Lange's What Makes a Scientific Explanation Distinctively Mathematical?
Here is a very simple example (inspired by Braine [1972], p. 144):The fact that twenty-three cannot be divided evenly by three explains why it is that Mother fails every time she tries to distribute exactly twenty-three strawberries evenly among her three children without cutting any (strawberries!)...[A] manipulability account of causal relations (such as Woodward's) says roughly that C is a cause of E exactly when systematic changes in E can be brought about by suitable interventions on C. Clearly, manipulation of the numbers of strawberries or children would bring about corresponding changes in the outcome of Mother's attempt....I agree with Mancosu, Lipton, and Kitcher that distinctively mathematical explanations in science are non-causal explanations... the categorical ground for a key's power to open a distant lock resides in the key's structure and the lock's structure, but the lock's structure is not a cause of the key's power (on pain of action at a distance on the cheap).
...not all explanations are causal...Spatiotemporal symmetries and the principle of relativity do not describe causal relations, even at a coarse-grained or abstract level. Rather, they impose constraints on the laws of nature...
Krugman's Ricardo's difficult idea describes it as "simple and compelling to those who understand it, but about which intelligent people somehow manage to get confused time and time again".
The Harvard-based "radical" or "heterodox" economist Stephen A. Marglin interviewed here regarding his book Raising Keynes.
The General Theory...never went beyond a simplified model [of interest] with only two financial assets - cash and bonds... central banks provide the missing link in Keynes's theory. Keynes disputed the mainstream view that a market equilibrium meant full employment; he believed instead that equilibrium would usually fall short of full employment. But if you don't have a theory for the level of the rate of interest, you don't have a theory of equilibrium at all. On its own terms Keynes' model was incomplete. We are thrown back on a non-market theory of interest rates!The final innovation of the book is more directly related to current policy proposals like the Green New Deal. This is the section of the book that deals with "functional" finance... the [GND] requisite tax increases would not be to balance the budget for the sake of balancing the budget (sound finance), but to keep aggregate demand in line with the productive capacity of the economy (functional finance)...in the context of full employment, it will require a reduction in demand somewhere along the line, starting with the reversal of Trump's tax "reform".
Economics claims to be primarily a descriptive subject, with a secondary normative function. But I think the relationship is exactly the opposite - economics is basically a normative discipline, and those normative beliefs shape our descriptive tools and analyses. That is not a criticism; it is rather a call for honesty about our political and moral presuppositions, and how they shape the way we think about the economy.
There's another aspect to economics, which I refer to in an earlier book, The Dismal Science, which is the constructive function. Economists not only describe and make judgments; they also try to shape the economy in the image of their economics. Given the predominance of the mainstream, this means shaping the economy in the image of a self-regulating market and curing any market failures by more and better markets. This would less pernicious if economists were upfront about their ideological presuppositions rather than claiming their economics is ideologically neutral.
Collaborated with Sen at one time.
Hein [2017] suggests there are five Post-Keynesian strands: fundamental Keynesian (via Hyman Minsky, GLS Shackle, and Sydney Weintraub); Kaleckian; Kaldorian; Sraffian/neo-Ricardian; and Institutionalist (via Veblen, Lerner, Galbraith). Aside from these, the Marxian approach is still an important heterodoxy, and see Harrod. Lavoie [2019] points out that:
[t]he main MMT authors - Randall Wray, Matt Forstater, Stephanie Bell-Kelton, Pavlina Tcherneva, Andrew Watts, Eric Tymoigne - were all tied to post-Keynesian economics from the very start. The only exceptions would be Scott Fullwiler, who came from the Institutionalist tradition, and William Mitchell, who was closer to the Marxian traditionThe stock-flow consistent approach is at the heart of post-Keynesian economics since the mid-1990s, and it was a critical contribution of Godley and Cripps [Macroeconomics (1983); Lavoie was a student of Godley].
The latter gives the macroeconomic balance sheet, with K the fixed capital stock,the only asset that belongs to firms and W, the financial wealth stock, which is ultimately attributed to households.
| Households | Firms | Government | Foreign | Σ | |
|---|---|---|---|---|---|
| Public Debt | +W | -W1 | -WE | 0 | |
| Fixed Capital | +K | K | |||
| Σ | +W | +K | -W1 | -WE | K |
Alex Williams comments in a review [he is a Levy Institute graduate, a post-Keynesian centre, the home of Randall Wray]:
...the macroeconomic detective is armed with the knowledge that balance sheets always balance. This simple insight, that every transaction has two sides, means that there are certain aggregate relationships between transactions that must obtain for the world economy. Knowing this, it's possible to chase actors across seemingly unrelated balance sheets to find where the system as a whole was forced to balance. From here, the skillful economist can identify the long-run tendencies that a given balance is likely to create. (Wynne Godley famously predicted the Global Financial Crisis in just this way, following US mortgage debt around the world and back.)
Specifically, on Michal Kalecki [Hein 2017]:
Kalecki's (1939) principle of effective demand, as an earlier alternative to Keynes's (1936) approach, has the profit share (or the wage share) as a determinant of the multiplier effect of short-run exogenous expenditures (investment, government expenditures or exports), together with the propensity to save out of profits...[ie] wage-led growth...a major contribution...has been his theory of distribution based on mark-up price setting of firms in incompletely competitive markets.Gross profits net of taxes = Gross investment + Capitalists' consumption + Government budget deficit + Export surplus - Workers' saving
Again from Alex Williams:
Pettis and Klein [2020] note in passing that the era of scarcity as the fundamental economic fact has ended, and in its place is the fundamental fact of demand. We have known since Keynes that the limiting factor in developed economies is not a scarcity of resources, but rather a scarcity of demand for finished output.
Hein is a little more circumspect, noting that the empirical support for a wage-led growth and demand are mixed, even within the same country, probably reflecting the analysis, viz short-term (more profit-led) v. longer-term scale (more wage led) effects.
Post-Keynesian-Kalecki models can include investment plans of firms
that can allow saving and investment to equalize at equilibrium,
so that
g = S/K,
remains true, where
K is the amount of homogenous capital, S the total
savings in the economy, and g the growth rate of capital. Dutt
[
2016] presents a particular PKK model with two groups "the top and the
rest", which allows "financialization and the increasing importance and
income of top managers (or CEOs and their allies and close subordinates)
[to] explain both lower rates of growth, and rising wealth and income
inequality especially with deficient aggregate demand".
Recall that Piketty's main interest is the historical relation between g (growth of the whole economy) and r, the rate of return on capital, see for example, Piketty [2015]:
In a representative-agent framework, what r > g means is simply that in steady-state each family only needs to reinvest a fraction g/r of its capital income in order to ensure that its capital stock will grow at the same rate g as the size of the economy, and the family can then consume a fraction 1-g/r. For example, if r = 5 percent and g = 1 percent, then each family will reinvest 20 percent of its capital income and can consume 80 percent.
NCM = New Consensus/Classical Macroeconomics [Clarida,Gali,Gertler; Goodfriend, King] - the current model underlying central bank interest as major lever.
More from Hein:
In a post-Keynesian macroeconomic policy mix, fiscal policies have a major impact on economic activity and the distribution of disposable income, and should thus actively take care of real stabilisation of the economy in the short and the long run, using government expenditures and taxation as tools without any autonomous government deficit targets. It thus means to follow a functional finance approach in the tradition of Lerner (1943) (Arestis/Sawyer 2003, 2004b, Setterfield 2009b). Potential limits to government debt in this kind of approach are a matter of controversy between those sympathetic to neo-chartalism and functional finance - what is now called "modern money theory" (MMT) (Wray 2012) - and the critics of such an approach (Palley 2015b). The relevance of government debt limits will depend on the precise institutional link between the government and the central bank, the international acceptance of the national currency, whether private and public debt is denominated in the domestic currency and so on (Lavoie 2013). In particular, if central banks act as a lender of last resort for the government and guarantee government debt, and private agents thus do not have to fear the illiquidity or insolvency of the government, the level of government debt or government debt-income ratios should be of minor concern, as has been pointed by the proponents of MMT.
Added 20200828: Perry Anderson's review of Tooze's Crashed discusses Tooze's use of stock-flow consistency in analysing 2008, and Cedric Durand's criticism that Godley saw a key advantage of his approach being that it integrated the financial and real economies, and Tooze did not discuss that nexus.
An interest in technology as prosthesis (a la extended mind). The phenomenology of sensing the world once the intermediation becomes transparent, or is the experience. And the centrality of technology to being an anatomically modern human, "How Things Shape the Mind" [Malafouris 2013], and coevolution of the cognitive niche. Thinks an emphasis on the computational mind ignoring the accidents and constraints of matter is deceptive. Ihde points to Merleau-Ponty [1962]:
The blind man's stick has ceased to be an object for him, and is no longer perceived for itself; its point has become an area of sensitivity, extending the scope and active radius of touch, and providing a parallel to sight. In the exploration of things, the length of the stick does not enter expressly as a middle term: the blind man is rather aware of it through the position of objects than of the position of objects through it. The position of things is immediately given through the extent of the reach that carries him to it, which comprises, besides the arm's reach, the stick's range of action.
This skips to some extent the primary/secondary qualities distinction, ISTM. Sensuous aesthetics is of the primary.
Zeitgeist; DLD lead all. Started ~10:30. P1 crux harder than all of us remembered - looks like 2-3 holds have come off. Topped out ~15:00. Rapped in 2 60m falls, walking out at 17:00.
See also Hintikka's work on game-theoretic semantics.
The game-semantic model of computation (Abramsky, Hylan, McCluskey) gives "an intrinsic (i.e., without recourse to another model of computation), non-inductive, non-axiomatic [model], which is similar to classic Turing machines, yet beyond computation on natural numbers (which let us call classical computation), e.g., higher-order computation" [Yamada 2019]. Such higher-order computation is carried out in typed functional language eg PCF - "computation that may take (as an input) or produce (as an output) another computation...Existing models of higher-order computation are either syntactic (such as programming languages and nested sequential procedures), axiomatic (such as Kleene's schemata S1-S9) or extrinsic (i.e.,reduced to classical computation by numbering whose 'effective computability' is often left imprecise" [ Yamada 2017 ].
Yamada [2017] "point[s] out that computation is an intensional concept in the sense that not only what its result (or extension) is but also how it computes, i.e., its algorithm, matters in an informal sense. For instance, the programs double(succ(5)) and succ(succ(double(5))) clearly have the same value, namely 12, but different algorithms...these are extensionally equal but intensionally different. Let us call informally the equality of algorithms intensional equality of computation."
Reviewed with a solution by Bartha and Hitchcock [1999]
Introduced by John Leslie as a thought experiment relevant to his Doomsday Argument. Sequential trial where batches of 1, 9, 90, 900... individuals drawn from an countably infinite population will terminate with 1/36 probability per batch. There is a infinitesimally small probability of "90%" of the population losing, even though each individual has a 35/36 probability of surviving the episode they are sampled into.
Mentioned here:
...two goals that we have when we have beliefs - believing truths and avoiding errors. When we have a belief, it gives us a chance of being right, but it also runs the risk of being wrong. In constrast, when we withhold judgment on a proposition, we run no risk of being wrong, but we give ourselves no chance of being right.
Barnum and Wilce [2012] introduce post-classical probability theory as a very abstract approach developed for quantum mechanics. There is a finite test space X of all outcomes along with the covering of these by non-empty sets (the tests), and probability weights summing to unity for each test, with Ω the set of all weights for X. A quantum version has a Hilbert space. A probabilistic model is a structure {X,Ω}. Scandolo [2019] applies such to "the information-theoretic foundations of thermodynamics and statistical mechanics...in arbitrary physical theories...[G]eneral probabilistic theories identif[y] the two main ingredients of any physical theory to be its compositional structure (how to build experiments) and its probabilistic structure (how to assign probabilities to experimental observations)". Acceptable theories need to hew to four information-theoretic axioms, informally stated as follows: Causality - No signal can be sent from the future to the past. Purity Preservation - The composition of two pure transformations is a pure transformation. Pure Sharpness - Every system has at least one pure sharp observable. Purification - Every state can be modelled as the marginal of a pure state, uniquely up to local reversible transformations"
Weilenmann and Colbeck [2020]:
Information causality is a candidate principle for singling out quantum theory. Roughly speaking the principle is that sending n bits of classical information from one party to another cannot give the recipient access to more than n bits of previously unknown information regardless of any pre-shared resources the parties may have.
Pawlowski et al [2009], in the paper that introduces the concept:
[I]nformation causality holds true even if the quantum bits are transmitted provided that they are disentangled from the systems of the receiver. This follows from the Holevo bound, which limits information gain after transmission of m such qubits to m classical bits.
Playing with flang-7. This is the Ubuntu-packaged older flang - recall that LLVM 10.0.0 24 released March 2020). Some clang options do not work as advertised. Here is some interpreted (actually JIT, I think) Fortran:
> cat test.f90
program test
write(*,*) 'Hello from Fortran'
end program test
> flang -emit-llvm -c -o test.bc test.f90
> lli -version
LLVM (http://llvm.org/):
LLVM version 10.0.0
Optimized build.
Default target: x86_64-pc-linux-gnu
Host CPU: nehalem
> lli -entry-function=MAIN_ -load=/usr/lib/x86_64-linux-gnu/libflangrti.so -load=/usr/lib/x86_64-linux-gnu/libflang.so test.bc
Hello from Fortran
> llvm-bcanalyzer test.bc
Summary of test.bc:
Total size: 19104b/2388.00B/597W
Stream type: LLVM IR
# Toplevel Blocks: 4
> llvm-dis < test.bc | less
> llc test.bc -o test.s
Also see https://www.llvm.org/docs/BitCodeFormat.html
Saliba [2002] summarizes the refinements that Islamic astronomers made to the Ptolemaic model, following an avalanche of contradictory observations from the 9th century CE onwards (exaggerated equinoctal precession and obliquity of the ecliptic). In the 13th century, Urdi's lemma gave a solution for the planetary equant - "a deferent that moved uniformly in place around the axis of the centre of the planet", and al-Tusi the Tusi Couple, which in the steam engine is the sun-and-planet mechanism, converting linear to circular motion.
Ibn al-Shatir (born ~1305, muezzin and time-keeper in Damascus) uses both of these in a refined geocentric model. Copernicus uses same Tusi Couple, and his diagram is lettered in the same order as that in al-Shatir's MS. Abbud [1962] showed that the numerical agreement between al-Shatir and Copernicus is also extremely tight.
Swerdlow and Neugebauer [1984] argued that this implied Copernicus had seen a translation or summary of al-Shatir, maybe via the Jewish The Light of the World [Morrison 2017]. The 11th Century Alfonsine Tables, for example, are the work of Jewish astronomers in Spain.
Copernicus probably jumped from here to a geostatic heliocentric system for all the other planets and then a straight heliostatic system. Blasio [2014] argues that the parallels are just that, given that "Copernicus happily cites numerous earlier Islamic sources - and there is virtually no evidence that [the later Islamic astronomical works] were available to him". Nikgahm and Ragep [2019] think the agreement is not explainable this way, suggesting Copernicus had seen diagrams of al-Shatir's "individualistic" model for Mercury, the planet that has the most complex Ptolemaic model, and that had a large number of quite different previously published alternatives.
The authors [2005] of Natural History of Ashkenazi Intelligence. Cochran has a blog https://westhunt.wordpress.com/ discussing human evolutionary genetics. Cochran and Harpending [2009] argue for increasing recent selection on human genomes in their book The 10,000 Year Explosion.
From the end of Cochran's review of Plomin's Blueprint:
Indeed, social scientists have done such a terrible job that it's hard to see how the field can be repaired. They wanted the false results they got, and they still do. I'm sure their descendants will as well. Isn't heritability grand?
That is, Ji, Natarajan, Vidick, Wright and Yuen, who showed that "the complexity of approximating the quantum value of a non-local game G is equivalent to the complexity of the Halting problem."
A two-player non-local game is played between a verifier and two cooperating players named Alice and Bob who cannot communicate with each other once the game starts. During the game, the verifier samples a pair of questions (x,y) from a joint distribution μ, sends x to Alice and y to Bob, who respond with answers a and b respectively. The verifier accepts if and only if D(x,y,a,b)=1 for some predicate D. The quantum value of a non-local game G, denoted by wq(G), is defined to be the supremum of the verifier's acceptance probability over all possible finite dimensional quantum strategies of Alice and Bob for the game G.
Scott Aaronson: "To say it more simply, entangled provers can convince a polynomial-time verifier that an arbitrary Turing machine halts."
Recursively enumerable (or computably enumerable) is
...the simplest of all complexity classes, a language is in RE if there is some Turing machine M such that x is in L if and only if M on input x accepts. For x not in L, M on x can reject or run forever. The classic halting problem, the set of descriptions of Turing machines that halt on empty input, is RE-complete.MIP* is the set of things provable to a classically random polynomial-time verifier by two separated provers with an unlimited number of quantumly entangled qubits.
[An earlier] paper [by Natarajan and Wright] showed that quantum entanglement actually gets more, much more, than classical provers...[this] get[s] a much stronger and tight result, and...disproving the Connes' embedding conjecture.
interactive proofs [are] where the verifier verifies the correctness of a statement, by interacting with the prover and by using randomness. As was shown later by Lund, Fortnow, Karloff, Nisan, and Shamir, interactive proofs seem to be much more powerful than standard proofs, as every language in PSPACE can be verified efficiently via an interactive proof, whereas only languages in NP can be verified efficiently via a standard proof.[In] multi-prover interactive proofs (MIPs)...there are several [only need 2] provers that are proving a statement to a single polynomial time verifier, and the assumption is that these provers do not communicate with each other during the proof...These can be converted into probabilistically checkable proofs (PCPs).
A first attempt to formulate [Charles H] Bennett's idea is to say that the logical depth of S, LD(S) is the time it takes for the shortest program of S, S*, to produce S [the decompression time]...The concept of thermodynamic depth introduced by Seth Lloyd & Heinz Pagels (1988) is defined as "the amount of entropy produced during a state's actual evolution" [the difference between the system's coarse- and fine-grained entropy]. It is a first attempt to translate Bennett's idea in a more physical context.
Paint spots now mark a route that traverses a little more westerly and then ascends to below the razorback. Descent really busy. Guys climbing Stainless Climb (the grade 27 free climb), trying to free the third (roof) pitch of Anticlimb.
| Code | Description | Count |
|---|---|---|
| 1777 | Part of a multiple birth | 11846 |
| 2734 | Number of live births | 1.8 (mean) |
| O30 | ICD10 Multiple Gestation | 108 |
| 132212 | Date O30 first reported | 641 |
| O31 | ICD10 Complications of Multiple Gestation | 21 |
| 132214 | Date O31 first reported | 35 |
| O84 | ICD10 Multiple Delivery | 22 |
...it would do no good to settle for saying that it is simply a matter of different "language games". Were we to do so, our generosity would actually be a cover for extreme stinginess, since it is to language, but still not to being, that we would be entrusting the task of accounting for diversity.
After crossing the Gateway bridge, rode to Cannon Hill Plaza. From there, regained Bulimba Creek Bikeway at Wynnum Rd, and went right through to where SE Freeway crosses Logan Rd just South of Garden City. Back along SE Freeway bikeway, where new section at Gaza Rd is now open. OK Coffee at Beartown Coffee House, Gabba.
The BLUPF90 webpage is more up to date than the PDF manual, notably for SSGWAS analysis, where there is no mention of the snp_p_value option described by Aguilar et al [2019]. This paper confirms that EMMAX/GCTA type analyses are equivalent to GBLUP, and that a single-step SNP-BLUP analysis can incorporate ungenotyped but phenotyped individuals. The resulting estimates include individual prediction error variances for all SNP effect estimates in one calculation. To obtain this analysis, one uses renumf90 to prepare the data files, and then runs blupf90 and postGSf90 in sequence.
Simple univariate example using Sib-pair:
read bin dataset.bin.gz
unique_id sequential
set print 01110
set mis "0"
drop
write ssgwas.ped
undrop trait
set print 01000
set ple -2
file delete ssgwas.dat
out ssgwas.dat
print where trait ^= x
out
undrop
set loc typed aff
if (protyp > 0.9) then typed = y
write blupf90 ssgwas.geno typed
write map blupf90 ssgwas.map
#
# Call renumf90
#
file delete tt_renum.job
out tt_renum.job
echo DATAFILE
echo ssgwas.dat
echo TRAITS
echo 2
echo FIELDS_PASSED TO OUTPUT
echo 1
echo WEIGHT(S)
echo
echo RESIDUAL VARIANCE
echo 0.5
echo EFFECT
echo 1 cross alpha
echo RANDOM
echo animal
echo FILE
echo ssgwas.ped
echo FILE_POS
echo 1 2 3 0 0
echo SNP_FILE
echo ssgwas.geno
echo (CO)VARIANCES
echo 0.50
echo OPTION map_file ssgwas.map
echo OPTION snp_p_value
out
$ renumf90 tt_renum.job
$ preGSf90 renf90.par
$ blupf90 renf90.par
$ postGSf90 renf90.par
The postGSf90 output file chrsnp_pval contains columns {trait, effect, -log10_P, SNP, chromosome, position} [see here].
Linquist et al [2020] summarize the nosology of "selected effect" functions and "causal role" functions (latter is any downstream effect). Also discuss Constructive Neutral Evolution [Stoltzfus 1999] where epistasis via effects such as buffering can allow a "neutral evolutionary ratchet", that is, buffering allows occult variation that then makes buffering compulsory ( "irremediable complexity"). Maynard Smith and Szathmary invoke this as the contingent irreversibility of major evolutionary transitions eg multicellularity. [via Larry Moran].
Belt, lubricant, locktab out of position are usual problems.
Try:
Unplug the unit for an hour press and hold eject button while plugging in invert the unit and quickly unplug when mechanical noise starts try ejecting when returned upright or while inverted eject and unplug just before starts moving
Latter was effective in this case.
Brielmann and Pelli [2017] find that "pleasure amplitude increases linearly with the feeling of beauty, To test Kant's claim of a need for thought, we reduce cognitive capacity by adding a "2-back" task. This added task greatly reduces the beauty and pleasure experienced from stimuli that otherwise produce strong pleasure, and spares that of less-pleasant stimuli. We also find that strong pleasure is always beautiful, whether produced reliably by beautiful stimuli, or just occasionally by sensuous stimuli. In sum, we confirm Kant's claim that only the pleasure associated with feeling beauty requires thought and disconfirm his claim that sensuous pleasures cannot be beautiful."
Luoto [2017] counter that "visually pleasing stimuli can cause affective and sexual responses even without conscious awareness [posing a] challenge to their argument (Ponseti and Bosinski, 2010; Gillath and Collins, 2016; cf. Chatterjee et al., 2009). The counterargument that thought is not a prerequisite for an affective response to visual stimuli is also supported by the finding that visual exposure to faces from out-group ethnical groups can elicit interracial affective bias outside conscious awareness (Yuan et al., 2017)...there are integrative processes that can occur outside of conscious awareness (Mudrik et al., 2014)... [A]n aesthetic judgment of beauty is firmly grounded in sensory processes (Jacobsen, 2006), and there are no empirical grounds to cleave visually mediated appreciation of beauty from sensory pleasures. Beauty can be amplified by cognitive processes (Vessel et al., 2012)¿such as integration with novel associations, integration over higher semantic levels, or integration over multiple modalities (Mudrik et al., 2014) - processes which can be particularly important for the experience of art (Nadal, 2013)."
Samuel K. Cohn Jr [2006] in Lust for Liberty: The Politics of Social Revolt in Medieval Europe, 1200-1425: Italy, France, and Flanders claims:
The first outright victory of this lower stratum in Italy appears in Bologna in 1289, when "the people without underpants" (servants, apprentices, "and others of a lower sort") booted Bologna's highest officer, the lord podesta, out of office and led him personally to the city gates. No retaliation followed, either from the city government, the forces of the podesta, or any faction of the ruling elites.I think this looks like a misunderstanding of the local government of that time in Bologna. Carniello [2002]:
Bologna's guild-based government from 1282 to 1292 offers an important case for the study of Italian popular reform movements...Rolandino Passaggeri, renowned Bolognese master of notarial arts, authored the [anti-Magnate] Sacred Ordinances, launching the popular government, and reorganised the notaries' guild as part of the reform initiative... Sometime before 1243, when conflicts turned to violence in the streets of the town, magnate families had already coalesced into two factions, the Guelf (called the pars Ieremiensium after the Geremei) and the Ghibelline (called the pars Lambertaciorum after the Lambertazzi)...The popular movement's response to magnate violence was the establishment of its own political structure, the populus, which was completed in 1255 with the establishment of the captaincy of the Popolo. The Ghibelline magnates [were] expelled in 1274 after the civil war, when the Ferrarans came in support of the Geremei... The Geremei, far from enjoying the fruits of victory in Bologna, followed their enemies, took their own crippling blow in battle against the combined Ghibelline forces in 1275, and had to turn to Charles of Anjou in 1276 for military assistance...The Geremei were not occupied with ruling Bologna, much less governing the town and working to free the streets from violence...[After being returned by papal intervention in 1279, the violence also returned and the] Lambertazzi expulsion had to be repeated...
The new government, the Popolo, comprised ministers from the Guilds, popular arms societies (militias), and magnates, but only in a personal capacity. A list of magnate family members had to pay a good behaviour bond, and pay restitution for the civil war. Serfdom was abolished.
The episode in 1289 may or may not be the Tumult of the Fullers. The Podestà is an outsider who served as chief magistrate hired by the city, and rotated out after each six month period. Corso Donati was a leader of the Florentine Black Guelph faction (related to Dante and mentioned in the Divine Comedy), and was Podestà of Bologna in July-December 1288. For 1289, they were Antonio de Fixiraga and Zachanus de Zachanis.
I thought I would put together a few more references here:
First, from Richard Pettigrew's remarks about Bayesian updating
[Re the] Bayesian norm of updating [,] [s]ome pay attention to the pragmatic costs of updating any other way (Brown 1976; Lewis 1999); some pay attention to the epistemic costs, which are spelled out in terms of the accuracy of the credences that result from the updating plans (Greaves & Wallace 2006; Briggs & Pettigrew 2018); others show that updating as the Bayesian requires, and only updating in that way, preserves as much as possible about the prior credences while still respecting the new evidence (Diaconis & Zabell 1982; Dietrich, List, and Bradley 2016). And then there are the symmetry arguments...
I can immediately see Darwinian(-like) arguments along the lines of those for success semantics. Poor reasoners, that is those who fail to follow appropriate norms of reason, will be underrepresented in the population, and those who can recognize that there are such norms (a type of "reflective knowledge") should have a further advantage. https://www.frontiersin.org/articles/10.3389/fpsyg.2018.01291/full
Rico the collie dog had a vocabulary of 200 words and would return an unfamiliar object from among a collection of familiar objects on command if asked for it by a novel word. I doubt he was up for polysemy.
One of the greatest disasters that befell twentieth-century analytic philosophy was Quine's (1953) rejection of the distinction between analytic and synthetic truths as an "untenable dualism"...the terms "analytic" and "synthetic" are most unfortunate from a historical point of view. What is meant is in fact a distinction between conceptual and factual information. Quine is right in effect pointing out that one cannot tell from a person's behavior whether the information he is relying on is factual or conceptual...[but this] does not mean that one cannot define the distinction by some other means. Quine's way of thinking, and that of many other contemporary philosophers, is hence predicated on a distinction between logical and nonlogical constants.The first observation that can be made here is that nonlogical analytical truths sometimes turn out to be logical ones when their structure is analyzed properly. In his Tractatus, Wittgenstein apparently assumed that this can be done for all conceptual truths. Within his truth-functional logic this amounted to the independence of atomic (elementary) propositions of each other. He ran into difficulties, however, in connection with color concepts, and was ultimately led to change his entire philosophy because of such difficulties. Wittgenstein despaired too soon, however, for at least in his paradigm case, the conceptual incompatibility of color terms can be turned into a logical truth simply by conceptualizing the concept of color as a function mapping points in a visual space into color space. (See Hintikka and Hintikka 1986, pp. 123-132) This is an instructive example of how nonlogical but "analytic" truths can be interpreted as logical ones. Then they are uninformative ("tautological") by the same token and in the same sense as logical truths.
He then introduces his Game-Theoretic Semantics (GTS) and independence-friendly (IF) alias hyperclassical logics.
The KK thesis in epistemic logic is that if an agent knows X, then they know that they know X. GTS can allow for imperfect memory and finite processing power ("liberated from classical linearity and perfect information").
Via Brighton, Thomas St Clontarf to Elizabeth Ave and Kipparing. Back via Petrie, Leitch's Crossing, Old Northern Rd (65 k). Notes at bora ring at Nudgee Waterhole mention Kippa Ring is a bora ring for initiation of boys (kippas).
| Compiler | Code | Time | Comment |
|---|---|---|---|
| gfortran 9.2.1 | rs17293443.in | 16m31.097s | -O2, openmp |
| sunf95 12.6 | rs17293443.in | 17m23.626s | -O3, openmp |
| Flang 7.0.1 | rs17293443.in | 10m56.857s | -O2 |
| Compiler | Code | Time | Comment |
|---|---|---|---|
| gfortran 9.2.1 | weil.in ae | 0m4.244s | |
| sunf95 12.3 | weil.in ae | 0m3.916s | |
| Flang 7.0.1 | weil.in ae | 0m12.290s | |
| gfortran 9.2.1 | weil.in ce | 0m14.624s | |
| sunf95 12.3 | weil.in ce | 0m13.531s | |
| Flang 7.0.1 | weil.in ce | 0m39.146s |
Flang and gfortran, it turned out, differ in treatment of signed zeros by sign().
Song from The Velvet Underground [1969] - the first post-Cale album. It has simultaneous recitations from Reed and Morrison of different sets of lyrics (one twice as fast) over organ and guitars, interspersed with Maureen Tucker and Doug Yule singing a more lyrical chorus in lagged counterpoint. Reminds me of Jefferson Airplane/West Coast (raga) crossing Steve Reich (especially the last two minutes).
Ebriety \E*bri"e*ty\, n.; pl. {Ebrieties}. [L. ebrietas, from.
ebrius intoxicated: cf. F. ['e]bri['e]te. Cf. {So?er}.]
Drunkenness; intoxication by spirituous liquors; inebriety.
"Ruinous ebriety." --Cowper.
Semiotician, phonologist (eg classified childhood aphasias) and linguist (Russia, Czechoskovakia, US) (and Formalist) whose six functions of language are:
| referential |
| aesthetic (materiality of means of communication) |
| emotive |
| conative (rhetorical, persuasive) |
| phatic |
| metalingual (reflexive) |
These might be compared to the Peircian classification of relationships between sign and signified: Icon, Index, Symbol. Note that any given pairing may be multimodal (as above). Jakobson argued that the aesthetic is an additional type.
The Russian Formalists (eg cognitive estrangement) see poetry as disautomatizing and refreshing expressions, images and themes [Winner 1987], so Jakobson [1960] summarises the aesthetic function as "orientation toward the utterance", using "self-valuable" words, visual or acoustic materials. Prague Linguistic Circle and Prague School aesthetics extends this, the work of art as a structure, and art as a sign. Jakobson saw the aesthetic function as autonomous to the other functions - "introversive semiosis" with minimal or polysemous cultural references.
Influenced Levi-Strauss. Caton [1987] is a review of his contributions:
classification of phonemes (consonants and vowels) using 12 binary features of spectral analysis (compact, grave, strident...), which then give dimensions for analysis of morphemes and further up the hierarchy, phrases.
a teleological view of language, where reference is not the primary goal, so "the sound system...cannot be analysed without taking into account the purpose which that system serves..".
Contextual meanings (Grundbedeutung) "cannot be acknowledged without the existence of [relevant] invariant meanings (Gesamtbedeutung). These correspond to semantics and pragmatics, and the influence of behaviourism split these apart in US linguistics. In the example of "shifters", the pronoun "I" has an invariant meaning (the addressor of the speech act), and a contextual Indicial, "the utterer is existentially related to his utterance [so "I" indexes him]" - deixis essentially starts as a field of study following Jakobson's 1957 paper. Jakobson also worked on iconicity of grammar (see below).
Parallelism is a formal aesthetic process eg in poetry parallel the rhyme of words with parallels in semantic category eg "shame" and "blame" as moral judgments and as nouns. Similarly, play off. This extends to grammatical structure, eg order inversions. He published several very close readings of poems eg Shakespearian sonnets 127, 129. Wallot and Menninghaus experimentally test effects of this on semantic processing:
In general, multiparallelistic sentences and texts are linguistic analogues to multilayered structures of symmetry, repetition, and variation that are well established as core features of the aesthetic appeal of both music and visual objects.[...] Roman Jakobson stipulated: "Ambiguity is an intrinsic, inalienable feature in any self-focused message".
Umberto Eco [1979] introduces ostention as a semiotic process - the signified of the ostended object is its class. Barry comments in the context of a semiotic analysis of one abstract painting that the significand might be intrinsically hard to comprehend (eg space and time).
Charles Morris is the logical empiricist (associate editor of the United Encyclopedia of Unified Science) who tried to present a semiotics of aesthetics esp literature, with the aesthetic sign as "an icon whose designatum is a value" [Steiner 1979]. He was not popular with the New Critics.
The Self-Destructing Modules Behind Revolutionary 1956 Soundtrack of Forbidden Planet
Rather than just tape effects/musique concrete, far more complex.
These are needed to assess the significance of the SKAT statistic, but applications are wide.
Chen and Lumley [2019] compare algorithms, and find the saddle-point approximation of Kuonen [1999, 2001] well behaved. There is an implementation in the R survey package.
Gold standard methods are those of Davies, Imhof, an implementation by Farebrother), and Rice [1980].
SPEN (gene spans ~100 kbp on 1p36) is the Xist-binding protein that is essential for X chromosome inactivation [reviewed by Trotman & Calabrese, 2020]. SPEN is suggested to bind active promotors and enhancers of target genes, complexed with Xist and then HDAC3 and NuRD. gnomAD records 4 (expect 129) pLoF variants, not increased in melanoma.
We use administrative data on Swedish lottery players to estimate the causal impact of substantial wealth shocks on players' own health and their children's health and developmental outcomes...Overall, our findings suggest that in affluent countries with extensive social safety nets, causal effects of wealth are not a major source of the wealth-mortality gradients, nor of the observed relationships between child developmental outcomes and household income.
Cited by Hill et al [2019] "Genome-wide analysis identifies molecular systems and 149 genetic loci associated with income".
There were three main findings: (1) DZ twin GPS differences predicted DZ differences in height, BMI, intelligence, educational achievement, and ADHD symptoms; (2) target and cross-trait analyses indicated that GPS prediction estimates for cognitive traits (intelligence and educational achievement) were on average 60% greater between families than within families, but this was not the case for non-cognitive traits; and (3) much of this within- and between-family difference for cognitive traits disappeared after controlling for family socio-economic status (SES), suggesting that SES is a major source of between-family prediction through rGE mechanisms...[P]revious within-family analyses have revealed substantial reductions in individual SNP effect sizes. For example, there was an effect size attenuation of ~40% compared to between-family associations in the most recent GWA study on educational attainment. Most of this reduction has been attributed to prGE; no similar deflation of effect sizes was found for height, indicating that prGE is not likely at play. A novel method relying on closely and distantly related individuals, and that is applied to very large populations, detected a similar reduction of SNP-heritability estimates of educational achievement (~40%).
Regression predicting SES from polygenic risk scores [Table S12]
| GPS | beta.B | L.CI.B | U.CI.B | P.B |
|---|---|---|---|---|
| ADHD | -0.168 | -0.214 | -0.13 | 1.32e-28 |
| BMI | -0.19 | -0.216 | -0.134 | 3.80e-38 |
| EA | 0.433 | 0.409 | 0.479 | 2.35e-204 |
| Height | 0.069 | 0.02 | 0.102 | 6.51e-06 |
| IQ | 0.231 | 0.213 | 0.291 | 4.28e-55 |
| Neurot | -0.06 | -0.129 | -0.043 | 5.35e-05 |
| SCZ | 0.04 | -0.015 | 0.066 | 0.008 |
| Health | 0.302 | 0.238 | 0.318 | 3.86e-89 |
Kong et al [2018] similarly quantify the effects of "genetic nurture" in Iceland: the polygenic score computed for the nontransmitted alleles of 21,637 probands with at least one parent genotyped has an estimated effect on the educational attainment of the proband that is 29.9% (P=1.6×10-14) of that of the transmitted polygenic score. Specifically, transmitted alleles: θT=0.223 with R2=5%, and θNT=0.067 with R2=2.5%. Adjusting for parental educational attainment halves the effect of the nontransmitted allele EA PRS on offspring EA. There was no significant difference between paternal and maternal contributions.
Cheesman et al [2019] found in a comparison of 6311 adopted versus nonadopted children in UKBB that "polygenic scores are twice as predictive of years of education in non-adopted individuals compared to adoptees (R2= 0.074 vs 0.037, difference test p= 8.23 × 10¿24).
In passing, Cox et al [2019] in UKBB (N=7201) found that "[t]he association between (age- and sex- corrected) total brain volume and a latent factor of general intelligence is r=0.276, 95% CI 0.252-0.300... largest brain regional correlates of g were volumes of the insula, frontal, anterior/superior and medial temporal, posterior and paracingulate, lateral occipital cortices, thalamic volume, and the white matter microstructure of thalamic and association fibres, and of the forceps minor."
Tian, Browning and Browning [2019] give 1.3×10-8 mutations per base pair per meiosis with a 95% confidence interval of 1.0×10-8, 1.6×10-8.
My postings in the discussion of John Wilkin's thoughtful essay Is racism Christian? (qv) include one on Sweet [1997], who argues Christian attitudes to black Africans were transmitted from the Moors:
By the ninth century, Muslims were making distinctions between black and white slaves...The white mamnuk commanded a higher price than the black 'abd because he could bring a substantial Christian ransom or be exchanged for a Muslim captive. The differing treatment of white and black slaves reflected their relative worth. The mamnuk was viewed as an investment to protect,.. [w]herever there was back-breaking work to be done in the Arab world, black slaves were made to do it...[f]rom ninth-century Iraqi land reclamation projects to fourteenth-century Saharan salt and copper mines...White slaves were...usually household servants....In the eleventh century,Toledo historian Sd'id al-Andalusi wrote [pretty much as per Aristotle on natural slavery]:.
For those peoples...who live near and beyond the equinoctal line to the limit of the inhabited world in the south, the long presence of the sun at the zenith makes the air hot and the atmosphere thin. Because of this their temperaments become hot and their humors fiery, their color black and their hair woolly. Thus, they lack self-control and steadiness of mind and are overcome by fickleness, foolishness, and ignorance. Such are the blacks, who live at the extremity of the land of Ethiopia, the Nubians, the Zanj and the like.
From an initial differentiation on ransom price, Sweet argues Iberian Muslims moved rapidly to an Aristotelian climate justification as well as the religious "Race of Ham" argument about why slavery was acceptable, even for some black fellow Muslims:
Islamic interpretations of Noah's curse varied, but a tenth-century Persian historian, Tabari, presented a typically racial response "Ham begot all blacks and people with crinkly hair...Noah put a curse on Ham, according to which the hair of his descendants would not extend over their ears and they would be enslaved wherever they were encountered."
Zanella et al [2019] report that BAZB1 (alias Williams syndrome transcription factor WSTF) patterns the human face via effects on neural crest precursors, and that regulatory mutations in downstream target genes seem more common in modern humans than in Neanderthals and Denisovans.
Wilkins et al [2014] previously hypothesized that "all the facets of the domestication syndrome can be traced to mild neural crest cell deficits". Zanella et al point out that "[a]mong the genes downstream of BAZ1B... uncovered in this study, FOXP2, ROBO1, and ROBO2 have long been implicated in brain wiring processes critical for vocal learning in several species, including humans, and will warrant further mechanistic dissection in light of the distinctive linguistic profile of Williams-Beuren [7q11-23 deletion or duplication] syndrome individuals". WBS is characterised by facial reduction and retraction, pronounced friendliness, and reduced reactive aggression, and "[s]tructural variants in WBS genes, for example in the case of GTF2I and its paralogs, have been shown to underlie stereotypical hypersociability in domestic dogs and foxes".
De dicto: Necessarily, some x is such that it is A
De re: Some x is such that it is necessarily A
De se: Of oneself, but quasi-indexed (and see logophoricity)
Is a first generation i7 much different from a Xeon (Sky Lake)?
| tinonee | i7 920 @ 2.67 GHz (8 GB RAM) |
| hpcnode063 | Xeon Gold 6242 @ 2.8GHz (32 GB RAM) |
| Task | tinonee | hpcnode063 |
|---|---|---|
| Sib-pair test suite | 5 s | oopsie! [1] |
| Compiling sib-pair | 72 s | 74 s |
| simulation with lamp [2] | 64.8 s | 40.7 s |
[1] "sp-00001505.kin" Fortran runtime error: Reading more data than the record size (RECL)
[2] Sib-pair script powersim2.in
An overview is at https://datascience.cancer.gov/data-commons/cloud-resources
https://www.cancergenomicscloud.org/
The TCGA PanCanAtlas is hosted at https://isb-cgc.appspot.com/, but ISB-CGC offers access to all of TCGA, as well as several other datasets. The link to the PanCanAtlas in Huang et al [2018] still points to an uncompleted page. But see https://github.com/ding-lab/PanCanAtlasGermline/blob/master/README.md.
A glossy overview of PanCanAtlas publications:
https://www.cell.com/pb-assets/consortium/PanCancerAtlas/PanCani3/index.html
gdc-client download -m ~/Downloads/gdc_manifest_20191113_012256.txt -t ~/Downloads/gdc-user-token.2019-11-13T01_25_27.008Z.txt
NIH bibliometric tool: https://icite.od.nih.gov/analysis. The Relative Citation Ratio is cites/year v. NIH funded papers in same field and year. They also give an NIH percentile, relative to NIH funded papers.
From formal logic to formal ontology: The new dual paradigm in natural sciences .
The possibilist modal logic KD45 is an optimal candidate for a formal ontology of Quantum Field Theory, that gives a semantic information based on the Boltzmann-Schrodinger notion, where
the macroscopic "ordered state", of which information measurement corresponds to the variation of the density distribution of the so-called "Nambu-Goldstone bosons" (NGB) - "phonons" in the crystal state of matter, "magnons" in the ferromagnetic state, DWQ [Dipole Wave Quanta] in the living state of matter....Dipole Wave Quanta -- corresponding, at the mesoscopic level, to the long-range correlation waves observed in brain dynamics -- depends on the triggering action of the external stimulus for the symmetry breakdown of the quantum vacuum of the corresponding brain state.
The International Research Area on Foundations of the Sciences [https://irafs.org/ is based at the Pontifical Lateran University.
In case any economic historian has been asleep or on Mars for the past three years, you may want to know that the economics-of-slavery culture wars have broken out again. Though only a pale shadow of the dust-up we had back in the 1970s, the aggressive assertions of the "new history of capitalism" regarding the centrality of slavery for U.S. economic development, and critiques of this work by economic historians, have generated much commotion in academic circles,
Beckert & Rockman Slavery's Capitalism: A New History of American
Economic Development
Sven Beckert Empire of Cotton
Walter Johnson River of Dark Dreams
Edward Baptist The Half Has Never Been Told
After Gödel. Platonism and Rationalism in Mathematics and Logic [2011] joins up phenomenology to mathematics.
van Atten et al similarly discusses the influence of Husserl and Kant on Brouwer (and Weyl). For intuitionists, the reals are "incomplete", as they are "indefinitely proceedable sequence[s] of 'nested' intervals" rather than atomic "finished" points. Sequences (including the above) can be lawful or lawless, but even the lawless represent a sequence of free choices. This is why intuitionist logic eschews the Principle of the Excluded Middle -
consider a lawless sequence α of which so far the intial segment 1,2,3, has been generated, and the statement P='The number 4 occurs in α'...we cannot say that P OR not P holds...[and] extensional identity of choice sequences (i.e. having the same values at the same places) is not generally decidable...[A] choice sequence α can be taken as an argument of a total function because in that case the function assignment must be contructable from just a suitable initial segment of α.
Blackburne [2004] reviews the pragmatist-derived idea (Frank Ramsey, Jamie Whyte [1990]) that practical success follows from correct representation:
the truth condition of a belief [is] that condition that guarantees the success of desires based on that belief.
Blackburne suggests "guarantee" might be too strong, and that "aid" is more correct. He argues that this approach is not obligatorily teleosemantic. One of Papineau's comments is that success or satisfaction of a goal is itself a "representational notion", so how does this help the ontology of representation? And what about pragmatic success based on a misunderstanding or a fiction?
Kong et al [2018] find that in islet cells,
p15 abundance did not correlate with p16 and only marginally correlated with p14 and ANRIL. MTAP expression was marginally correlated with p14, p16, and ANRIL, but highly correlated with p15 expression. [r=0.68]
MTAP and p15 transcripts levels were not correlated with age, cf p16.
Recall Sangalli et al [2017] found the rs7023954*A allele to be more expressed than the G in fibroblasts, and that that haplotype was less methylated.
Worm [Dave DeLaney "not grimdark, but crapsack"] See https://tvtropes.org/pmwiki/pmwiki.php/Recap/Worm Taylor Varga Mauling Snarks ceruleanscrawling.wordpress.com
https://github.com/hmgu-itg/VCF-liftover
discusses the chain file format. The header for each chain gives the chromosome (fields 3 and 7), start of the current interval (field 6) and offset between maps (field 11 - field 6). The following lines of the chain give the end of the current interval (start + field 1), the new start, (start + field 1 + field 2), and the new offset (old_offset + field 3 - field 2).
For example:
chain 3231099988 chr22 50818468 + 16367188 50806138 chr22 51304566 + 16847850 51244566 23 19744 0 40 36 1 1
becomes:
chr22 16367188 16386932 480662 chr22 16386932 16386968 480702 chr22 16386969 16387000 480702
I have implemented this in Sib-pair, and all seems to work.
https://github.com/materialsintelligence/mat2vec
Here we show that materials science knowledge present in the published literature can be efficiently encoded as information-dense word embeddings (vector representations of words) without human labelling or supervision. Without any explicit insertion of chemical knowledge, these embeddings capture complex materials science concepts such as the underlying structure of the periodic table and structure-property relationships in materials. Furthermore, we demonstrate that an unsupervised method can recommend materials for functional applications several years before their discovery....To train the embeddings, we collected and processed approximately 3.3 million scientific abstracts published between 1922 and 2018 in more than 1,000 journals deemed likely to contain materials-related research, resulting in a vocabulary of approximately 500,000 words. We then applied the skip-gram variation of Word2vec, which is trained to predict context words that appear in the proximity of the target word as a means to learn the 200-dimensional embedding of that target word, to our text corpus...
We find that, even though no chemical information or interpretation is added to the algorithm, the obtained word embeddings behave consistently with chemical intuition when they are combined using various vector operations (projection, addition, subtraction). For example, many words in our corpus represent chemical compositions of materials, and the five materials most similar to LiCoO2 (a well-known lithium-ion cathode compound) can be determined through a dot product (projection) of normalized word embeddings...[For example, ]
ferromagnetic - NiFe + IrMn = antiferromagnetic
...For instance, CsAgGa2Se4 has high likelihood of appearing next to "chalcogenide", "band gap", "optoelectronic" and "photovoltaic applications": many good thermoelectrics are chalcogenides, the existence of a bandgap is crucial for the majority of thermoelectrics, and there is a large overlap between optoelectronic, photovoltaic and thermoelectric materials...
Various methods have been developed to improve power for detecting sparse alternatives in this situation. The Tippett's minimum p-value test (Tippett, 1931), the higher criticism test (Donoho and Jin, 2004), and the Berk-Jones test (Berk and Jones, 1979) are particularly popular
The Higher Criticism test is based on the distribution of the P-values as order statistics under a normalised empirical process,
HC*n = max(i=1..pcrit*n) n0.5 [ i/n-p(i) ]/ (p(i)(1-p(i))0.5
The test is constructed for the alternative hypothesis of a two-component mixture, and critical threshold approximately
(2 log log n)0.5
due to "Law of the Iterated Logarithm" [Robbins 1970]. Robbins and Siegmund [1970,1973,1974] produced a similar sequential test, the Power-One Test, which in this setting is used for the exponential hypothesis. The R GHC package is one implementation of Generalized Higher Criticism for GWAS data.
Sun et al [2019] discuss set-based hypothesis testing of a number of tests, where the Berk-Jones statistic is Zmax (assuming independence of tests),
...the maximum of a set of likelihood ratio tests performed on S(t) [number of test statistics greater than threshold t] at all observed test statistic magnitudes greater than or equal to the median observed magnitude.
The Sun et al version incorporates a correlation matrix - in their GWAS setting, the marker LD - to give an analytic P, which they claim comparable to GSEA, SKAT etc.
Liu and Xie transform and sum P-values as Cauchy-transformed - Pillai and Meng [2016] found that the sum of perfectly dependent Cauchy variables follows the same distribution as if they were i.i.d., and under more complex dependence patterns, the (heavy) tails are not greatly affected.
T = S wi tan { 1-0.5piπ } with the w's summing to unity.
Horsman et al [2014] suggest:
...the use of a physical system to predict the outcome of an abstract evolution.
as opposed to an experiment, where evolution of a physical system tests the adequacy of the abstract model.
Scott Aaronson suggests it is whenever you can get a wrong answer (contra pancomputationalism etc).
Millhouse [2019] considers the collection of functions mapping physical states to abstract machine states and those mapping inputs to the abstract machine to interventions in the physical system. The least (algorithmically) complex interpretation is best.
A physical system, P, implements a machine, M, to the extent that the simplest interpretation of P as M, argminI in IK(I), is simple relative to the complexity of M, K (M). I is simple relative to M to the extent that it minimizes K(I)/K(M).
A guitarist and singer/songwriter playing blues, jazz etc from a young age with father - a famous composition is Strawberry Letter 23.
It was often argued that "all rational probability judgments are countably additive, but de Finetti denies this" [ Nielsen et al 2019]
Consider a lottery consisting of countably many tickets, one for each integer. According to de Finetti, it should be open to a rational agent to consider such a lottery fair (de Finetti 1974). To do so, each ticket must be assigned equal probability. But if probabilities are countably additive, this is not possible.
I am still thinking about this. It seems to me that a nonstandard analysis is possible a la Edward Nelson, following his derivation of the ordinal and cardinal versions of the Borel-Cantelli theorem (Theorems 7.3 and 7.4 of Nelson [1987]). Ai...Av are the events where one of the first v tickets wins, where v is a nonstandard integer infinitely close to +infinity. The question is whether the probability of a win of any of the first v tickets converges or not. If a rational agent has accepted that this is a fair lottery, this should converge to 1, and the AN are independent. Then
Π(n=i..v) (1-Pr An) ≤ exp(-Σ(n=i..v) Pr An)
as 1-x ≤ exp(-x). From this, the sum Σ(n=i..v) Pr An will be infinitely close to zero, even though the product on the left approaches unity. It is now possible to assign infinitesimal probabilities to each ticket (Robinson's lemma is that if x1..xn are infinitesimals, then there is one nonstandard v where xn are infinitesimals for all n ≤ v. An infinitesimal x is smaller than 1/n for all n in standard N), and two infinitesimals are infinitely close to each other.
Pruss [2018] tries to argue that this type of infinitesimal probability gives rise to inconsistencies, eg comparing two infinite lotteries - one where there is one winning number to one where there are two winning numbers (his example is where winning is a coarsening - double and round). Then one would prefer the second, even though it is only "weakly greater". But being offered a finite number of tickets in multiple infinite lotteries still offers only an infinitesimal chance of winning. Some of these problems arise after conditioning on having already drawn a ticket, so that the "face value" is a standard natural number. Stipulating that the values can be negative fixes some problems. Further stipulating that it is irrational to care about infinitesimal differences also helps.
A related argument is that Bayesian rationality cannot be normative in the face of infinities and topological set concepts such as meagreness, as they can be epistemically immodest approaching the limit of a series on conditionalizations (Bayesian Orgulity).
The Omnific is a progressive instrumental band from Melbourne, Australia featuring two bass players and a drummer.
Include tapping a la stick or Warr. Usually add synth backing/melismata, often ending up with a sound that reminds me of Metheny's Orchestrion.
Tim Willis and The End are a 5-8 piece jazz-rock group formed in 2010, who play with "post-rock and minimalism".
The End are Jack Beeche on alto sax; Jon Crompton (Wangaratta Jazz Awards Runner Pp 2009) on alto-sax; John Felstead (University Medalist, Lee Barker, Cam McCalister) on tenor sax; Tim Willis (composer) on lead guitar; bassist Gareth Hill (Michelangelo & The Tin Star, Bob Sedegreen); and Nick Martyn (Whitesploitation) on drums.
The Leveller, "Free-born John", who was so disputatious that
If the World was emptied of all but John Lilburne, Lilburne would quarrel with John, and John with Lilburne.
He was imprisoned on multiple occasions, escaping execution for returning from exile by "successfully" arguing "that the prosecution was unable to prove he was the same John Lilburne who had been banished". At an earlier treason trial
...before the court of Star Chamber, he refused to take the oath. "It is this trial that has been cited by constitutional jurists and scholars in the United States of America as being the historical foundation of the Fifth Amendment to the United States Constitution. It is also cited within the 1966 majority opinion of Miranda v Arizona by the U.S. Supreme Court."[He]...refused to take an Oath to answer Interrogatories, saying it was the Oath ex Officio, and that no free-born English man ought to take it, not being bound by the Law to accuse himself, (whence ever after he was called Free-born John)...
Like HP and the Methods of Rationality:
https://tvtropes.org/pmwiki/pmwiki.php/Fanfic/Luminosity http://luminous.elcenia.com/
I enjoyed this - only a few dud notes. However, I have not read the originals.
Athena Andreadis recommends Minna Sundberg's Stand Still, Stay Silent webcomic
http://www.sssscomic.com/comic.php?page=1
Really good.
...performed genome-scale CRISPR knockout (GeCKO) screen. We mutagenised the HAP1 cells with the GeCKO v2 library, which targets 19,050 human genes with 123,411 unique guide sgRNA sequences [ Sanjana et al 2014], and then selected these knockout pools with a lethal concentration of [box jellyfish] venom for 14 days.
Previously, we used a Genome-scale CRISPR Knock-Out (GeCKO) library to identify loss-of-function mutations conferring vemurafenib resistance in a melanoma model [ Shalem et al 2014].
Politicial philosopher from communitarian and socialist traditions. Key book is Spheres of justice: A defense of pluralism and equality (1983).
From Joshua Cohen's review:
The central thesis of [his] theory of value is a version of communitarianism:Trappenberg [2014] summarises:(C) The subjects [bearers] of values are in the first instance political communities, and not the individual members of those communities.
(C1) The objects that are socially valued are different for different political communities.
(C2) Communities typically have pluralistic values. That is, they value a variety of social goods - for example, money, political power, education, free time, love - which are unordered, in that there is no ranking of their relative value.
The second main element of TCE fits this theory of value into an account of the justification of distributive norms. (N) Each of the heterogeneous goods in a society is associated with a correct distributive norm, and that distributive norm is contained in the socially shared understanding of that good.
[Unordered plurality entails] a set of distinct "spheres of justice," each with its own internal regulative principle. What justice then requires is the "autonomy" of these spheres.
[J]ustice takes a different shape in different societal spheres. In the sphere of education justice has to do with creating equal opportunities (in primary education) and with rewarding according to merit (in secondary education). In the sphere of money and commodities, justice takes the shape of free exchange. In the sphere of welfare, goods are distributed according to `socially recognized needs'. In the sphere of politics justice is about procedures: democratic elections, the will of the majority, gaining the public's favour and so on. Each sphere of justice has its own `internal moral logic'. A political community should be ordered in such a way that its spheres of justice can uphold their internal moral logic.If a political community manages to keep its spheres of justice apart, such a community accomplishes an ideal called `complex equality'. Simple egalitarians are egalitarians who abhor income inequalities as such. Complex egalitarians on the other hand can put up with quite a lot of inequality in the sphere of money and commodities as long as this inequality is confined within that particular sphere. Rich people should not be able to buy political power, love and friendship, better education for their children, or preferential treatment by doctors, judges and policemen. But if these conditions are fulfilled there is nothing wrong with richness per se.
Walzer's friendly critics, the mitigated pluralists, seem to share a certain dislike for...the principle of spherical autonomy. They try to improve Spheres of Justice by seeking or constructing `across spheres criteria',`overarching principles' `underlying notions that go beyond local autonomy' and `non sphere-specific considerations
Economist (1910-1993) with interests in General System Theory (9 levels of organisation), ecological and evolutionary economics, peace studies, ethics of economics and economics as a moral science. A student of Keynes.
"Keynes saw himself writing in the Cambridge tradition of economics as a 'moral science'"
Here, John Danaher analyses a paper by Neil Levy, on how consciousness is important to moral value. I wrote:
Yes, as an attempt at a reductio wrt phenomenal consciousness and Siewert's wager. P1a. One must be sentient to have some kind of moral standing. P1b. Minimally, one must be able to suffer. P2. Suffering requires phenomenal consciousness (PC). C1. PC is a key feature of personhood. P3. Vivid access consciousness (memory, imagination) is experientially equivalent to (indistinguishable from) PC for some humans, by all accounts. P4. For other humans, similar mental function is present but as a less vivid access consciousness. P5. Focal loss of the capacity for such experiences due to neurological diseases does occur, without effects on other facets of consciousness. (Affected individuals notice the effect, and find it annoying and a loss of function). C2: Access and phenomenal consciousness can be functionally interchangeable in particular domains between persons. I think this is stronger than for perceptual faculties eg cortical blindness (v. blindsight). C3: By a stepwise replacement model, it seems plausible to me that we can now posit being a person with no phenomenal experiences. Say, for example, this person accesses visual information about the world only via working or short term memory. Hey, isn't that all of us absent-minded people ;)? Maybe PC is incoherent? At the very least, C1 seems contestable. Consider suffering due to acute tissue injury, psychogenic pain, and anticipated pain, and types of consciousness in each case.
Fine [ 2014] proposes a new realist semantics crossing contructivist and Kripkean (1965) "condition-oriented" semantics. In Brouwer-Heyting-Kolmogorov semantics, "a construction establishes B .and. C if it is the combination of a construction that establishes B and a construction that establishes C", while for Kripke, "a state verifies [or forces] B .and. C if it verifies B and verifies C". For the constructivist, the semantics is exact, but in the latter case, one can add (extend) irrelevant content to the state without changing the verification: "the state of the ball being red and of its raining in Timbuktu will also verify that the ball is red" - in the case of possible worlds verifying X, then there are relevant and irrelevant facts, though tricky regarding negation.
In intuitionistic logic, the latter is less of a problem. Fine defines a conditional connection between states (of nature) s -> t, that is s "leads to" t, where s is an exact verifier of, say, B. Now logically equivalent statements may not be verified by the same states. For example, the logically equivalent statements p and p .or. (p .and. q). A state might exactly verify just (p .and. q) but not p. Truth of a statement requires exact verification by an actual state ie a fact.
Ciaredelli derived a very similar generalized or inquisitive semantics [2009,2011,2013], under the model that propositions are proposals, which are defined as the set of all maximal states (roughly independent states) that support the proposition. A proposition can be contradictory, thus inviting an informative response to which arm is correct.
Eight founder strains: A/J (A), C57BL/6J (B), 129S1/SvImJ (C),NOD/ShiLtJ (D), NZO/H1LtJ (E), CAST/EiJ (F), PWK/PhJ (G),and WSB/EiJ (H). The 3 wild-type founders are F-H. There are ~70 RI strains currently available, as interstrain allelic incompatibility has limited the number of viable lines.
The Mouse Universal Genotyping Array has gone through three generations: 7851 markers, MegaMUGA 77808 markers, and now the GigaMUGA (Infinium HD) with 143259 markers.
Broman et al [2019] describe the updated R/qtl2 package, which can handle "multiparent populations derived from more than two founder strains, such as the Collaborative Cross and Diversity Outbred mice". This requires SNP genotypes for individuals and founders, coded as 1 and 3 for the different homozygotes (heterozygotes=2 are ignored) in the CSV genotype file
Power calculations for Haley-Knott regression of CC strain means are presented by Keele et al [2019]. They note the QTL support interval is roughly 10 Mbp, after allowing for LD. For r replicates,
h2QTL(r=1) = h 2QTL / (h2QTL + h2strain + s2/r)
For example, a mapping experiment on strain means with QTL effect size h2QTL=0.3, h2strain=0.4, s2=0.3, and r=10, is equivalent to our simulation of a single-observation with no strain effect but QTL effect size h2QTL=0.41.
| QTL h2 | Power (50 strains) | ||||
|---|---|---|---|---|---|
| 1 Obs | 3 rep* | 5 rep* | 2 alleles | 3 alleles | 8 alleles |
| 0.3 | 0.125 | 0.079 | 0.105 | 0.118 | 0.116 |
| 0.35 | 0.152 | 0.097 | 0.194 | 0.207 | 0.261 |
| 0.4 | 0.182 | 0.118 | 0.298 | 0.335 | 0.383 |
| 0.45 | 0.214 | 0.141 | 0.456 | 0.467 | 0.539 |
| 0.5 | 0.250 | 0.167 | 0.620 | 0.630 | 0.712 |
Note that there are "inconsistencies on CC haplotype probability file in build 38 compared to build 37. The two errors are as follows: On chromosome 5, there are two markers (SAbGeoEUCOMM001 SAbGeoEUCOMM002) that should not be given genome position. On chromosome 13, there is a problem with the last set of approximately 80 markers that result in an inconsistent pattern of founder haplotype reconstruction. (Beginning near UNC23486670 CH13:118447298 to UNC23498758 CH13:119480991)" [http://csbio.unc.edu/CCstatus].
Anthropologist-cum-neuroscientist-cum-biosemiotician currently at UC Berkeley - key books are The Symbolic Species: The Coevolution of Language and the Brain and Incomplete Nature: The Emergence of Mind from Matter, the latter summarised in detail in its Wikipedia article.
In Incomplete Nature, he presents:
...three modes of system dynamics that are distinguished by their hierarchic (i.e., nested) dependencies and their reversals of spontaneous global dynamical tendencies to reach different kinds of stable end-states (or attractors), if they are provided with the required time to do so. These dynamical modes include homeodynamics (e.g., processes at or near thermodynamic equilibrium), morphodynamics (e.g., non-chaotic dissipating processes such as exemplified by self-organizing systems), and teleodynamics (e.g., self-preserving processes such as exemplified by living systems).[Deacon and Koutroufinis 2014]
On reference, thinks it can be formalized in a physical model,
...reference is made possible by the susceptibility of a given information medium to reflect the effect of work with respect to an extrinsic context, and that the sign of this effect - i.e. whether there is an increase or decrease of medium constraint - will depend on whether this work originates in the interpretive process or in its extrinsic physical context.
The HERVs are 8% of the human genome, and have contributed two important coding genes. Syncytin-1 (ERVWE1 7q21) induces syncytia, and is highly expressed in the syncytiotrophoblast. It may modulate the immune response in pregnancy. Syncytin-2 (6q24) expression is limited to the villous cytotrophoblasts, and its receptor MFSD2 is in the syncytiotrophoblasts.
Numerous examples of de novo emergence of genes from previously noncoding sequences are now recognised - excluding horizontal gene transfer and poor annotation. "the RNA-first model, in which the formation of an ORF occurs in a region that is already transcriptionally active, is supported by previous reports on de novo genes and by both pervasive transcription and pervasive translation" [Vakirlis et al 2018]. Carvunis et al [2012] considered a "continuum" model that there is a large pool of protogenes being transcribed at low levels, some of which are then selected.
Vakirlis et al [2018] screened yeast data to find 366 taxonomically restricted genes that represented likely true de novo events - 0.5-1% of loci. A point mutation giving rise to an ORF could be identified in 30 cases. In one example in the cod, the promotor region for an antifreeze glycoprotein became associated following a translocation of the microsatellite (9 bp tripeptide repeat). Understandably, de novo genes tend to be shorter, and "biosynthetic cost is also lower than those of noncoding sequences, in agreement with an intermediate stage from a noncoding to a coding state". They tend to be reverse orientation to 5' neighbours and arising in GC rich areas, given that AT-rich stop codons are less common.
Chen et al [2015] describe 64 hominoid-specific "motherless" or "orphan" protein coding genes that arose from GC-rich lncRNA genes. Of these, 43 encode human-specific proteins, and 21 encoding similar proteins in human and chimpanzee but not in rhesus macaque. Protein expression was confirmed by mass spectrometry. In the macaque, the orthologous loci for the lncRNA precursors of human de novo genes are not subjected to strong selective constraints, ie "the ancestor of de novo genes may not be particularly distinct in terms of functional importance before the proteins arise". The human de novo genes do show NS/S ratios consistent with selection. Several studies have shown that the protein structure tends to be "disordered", but can still be functional.
Ruiz-Orera et al [2015] report "over five thousand new multiexonic transcriptional events in human and/or chimpanzee that are not observed in" macaque or mouse. Most did not evidence purifying selection. In human large-scale RNA-seq, "unannotated genes represented 0.5-2% of the transcriptional cost depending on the tissue...The vast majority of de novo transcripts were expressed in testis (93.8-94.5%), as were transcripts from phylogenetically conserved genes. In contrast, in brain, liver and heart, transcripts from de novo genes were underrepresented when compared to transcripts from conserved genes." Further, the de novo genes were AT rich compared to older genes. Only 6 were found to produce protein.
Via http://mindblog.dericbownds.net/ Ferreri et al [2019] show levodopa increases musical pleasure, and risperidone diminishes it. And earlier, Blood and Zatorre showed musical "chills" to be associated "with increased blood flow in the ventral striatum, the amygdala, and other brain regions associated with emotion and reward".
| Enhancer sites | H3K4me1 enriched. Active H3K27ac | ||
| Stretch enhancers | stretches of enhancers > 3 kb | cell-type-specific identity | |
| Super enhancers | long stretches of H3K27ac active enhancer modifications | mark LCRs | |
| Broad domains | contiguous H3K4me3 promoter marks >4 kb | cell-type-specific identity/function | eg G6PC2 |
Roadmap Epigenomics Consortium [2015]:
The active states (associated with expressed genes) consist of active TSS-proximal promoter states (TssA, TssAFlnk), a transcribed state at the 5' and 3' end of genes showing both promoter and enhancer signatures (TxFlnk), actively-transcribed states (Tx, TxWk), enhancer states (Enh, EnhG), and a state associated with zinc finger protein genes (ZNF/Rpts). The inactive states consist of constitutive heterochromatin (Het), bivalent regulatory states (TssBiv, BivFlnk, EnhBiv), repressed Polycomb states (ReprPC, ReprPCWk), and a quiescent state (Quies) which covers on average 68% of each reference epigenome. Enhancer and promoter states cover approximately 5% of each reference epigenome on average, and show enrichment for evolutionarily-conserved non-coding regions41.
ChromHMM:
| Active TSS | 0.7% |
| Flanking active TSS | 0.5% |
| 3' 5' transcription | 0.1% |
| Strong transcription | 3.6% |
| Weak transcription | 11.6% |
| Genic enhancers | 0.4% |
| Enhancers | 2.8% |
| ZNF genes and repeats | 0.2% |
| Heterochromatin | 2.6% |
| Bivalent/poised TSS | 0.1% |
| Flanking Bivalent/poised TSS | 0.1% |
| Bivalent Enhancer | 0.1% |
| Repressed Polycomb | 1.2% |
| Weak Repressed Polycomb | 8.3% |
| Quiescent/Low | 67.8% |
Polycomb repressive complex-2 (PRC2, subunits EZH2, SUZ12, EED, RBBP4, AEBP2) is a histone methyltransferase required for epigenetic silencing - adds three methyl groups to lysine 27 of histone H3.
Finucane et al [2015] partitioning of heritability of 11 traits by functional class:
...breakpoints statistically associate with features including regions of high gene density, high GC content, and high repeat content... Here, we show that the distribution of rearrangements can be accurately explained as misrepaired breaks between open chromatin regions in non-coding regions that are brought into contact by the three-dimensional conformation of chromosomes in the nucleus, which also provides a direct explanation for their mechanism of occurrence.
For rice, 72 Mb DNAse sites, 389 Mb genome.
| Interval, | Rice genome |
|---|---|
| exon | 1.9% |
| 3'UTR | 1.4% |
| promoter | 55.4% |
| intron | 15.7% |
| intergenic | 5.7% |
https://github.com/deruncie/GridLMM , as the name implies, tests for marker association (and GxE) against a grid of random effects values. Authors argue performant for case of multiple REs.
Existing systemic therapies for AD: cyclosporine, azathioprine (higher side effects), methotrexate and mycophenolate mofetil ("blocks de novo guanine synthesis via the inhibition of inosine monophosphate dehydrogenase leading to impaired leucocyte proliferation").
Novel systemic therapies (~25 trials to date)
| Mepolizumab | monoclonal antibody to interleukin-4 (IL-4). |
| Omalizumab | anti IgE antibody. |
| Dupilumab | IL-4 receptor-alpha antagonist (IL-4/IL-13) |
| Lebrikizumab | anti IL-13 |
| tralokinumab | anti IL-13 |
| Ustekinumab | p40 subunit of IL-12 and IL-23 (Th1 and Th17) |
| Nemolizumab | anti IL-31 receptor A. |
| Fezakinumab | IL-22 antagonist. |
| Baricitinib | (JAK1 and JAK2 inhibitor. |
NICE has recommended dupilumab as an option for treating moderate-to-severe AD in adults where the disease has not responded to at least one conventional systemic therapy or where these are contraindicated or not tolerated.
Sikaflex marine & polyurethane based construction adhesive
310 ml for $24.26 at Bunnings.
Gear Aid Freesole (aka Aquaseal) urethane formula: K2 $17.95 for 28 g.
Original Gorilla, Barge Cement
Some microsatellites are still "not mapped to the assembly in the current Ensembl database", but are on UCSC B37.
Japanese studies (Nakashima et al 2010, Ogawa et al 2014) pointed to rs8032158 on chromosome 15 as being associated with keloid scarring. More recently, the risk genotype C/C been found to increase expression of the transcript variant (TV) 3 of NEDD4 in keloid tissue. A quick look at UKBB finds the C allele associated with "M72 fibroblastic disorders" (P=10-14) as well as decreased sitting height. The ICD10 classification M72 covers fasciitis (eosinophilic, necrotizing, pseudosarcomatous), fibromatosis, and Duyputrens.
Marneros [2019] comments:
Higher transcript levels of NEDD4-TV3 are associated with increased activation of NF-kB and STAT3 in keratinocytes and fibroblasts of keloid lesions when compared with normal skin in individuals without this genetic risk allele. Overexpression of NEDD4-TV3 in primary human keratinocytes increased NF-kB activity, whereas its knockdown had the opposite effect.
Indigenous Australian singer/songwriter from Delungra NSW...won Triple J's National Indigenous Music Awards competition in 2012.
Clair de Lune and Candle good.
Bamber [1975] showed that AUC = Pr(Y ≥ X), that is the probability that the value for a randomly sampled case ≥ that of a control. It pretty directly follows that the Mann-Whitney U is a nonparametric estimator of AUC, allowing asymptotic intervals that tend to be less accurate when AUC is high and N small.
Huang et al [Nat Gen 2017] published their LINSIGHT scores for all sites, which they claim improves predictions for noncoding sequence via inference of negative selection (similarly to fitCons).
Probability of being a CDKN2A carrier [Taylor et al 2019]:
alogit(1.99+0.92*N_primaries+0.4*N_CMrelatives-2.11*log(Dx_age)
That is, sporadic case in 60 y.o., Pr=0.003; affected relative, two primaries, 20 y.o., Pr=0.11.
[F]irst large-scale, unbiased genetic study of historical EPP rates in a Western European human population based on combining Y-chromosomal data to infer genetic patrilineages with genealogical and surname data, which reflect known historical presumed paternity. Using two independent methods, we estimate that over the last few centuries, EPP rates in Flanders (Belgium) were only around 1-2% per generation.
Based on the analysis of 68 representative genealogical pairs, separated by a total of 1013 fertilization events, we estimated that the historical EPP rate for the Dutch population over the last 400 years was 0.96% per generation (95% confidence interval 0.46%-1.76%).
Here we...investigate 1273 conceptions over a period of 330 years in 23 families of the Afrikaner population in South Africa. We use haplotype frequency and diversity and coalescent simulations to show that the male population did not undergo a severe bottleneck and that paternity exclusion rates are high for this population. The rate of cuckoldry in this Western population was 0.9% (95% confidence interval 0.4-1.5%)...
Jiang et al [2019] use LDSC
| - | h2 | rg | |||||
|---|---|---|---|---|---|---|---|
| Breast | .12 | 1 | |||||
| CRC | .10 | .11 | 1 | ||||
| Head&Neck | .10 | .15 | - | 1 | |||
| Lung | .08 | - | .28 | .57 | 1 | ||
| Ovarian | .04 | - | - | - | - | 1 | |
| Prostate | .18 | .07 | .11 | .15 | - | - | - |
The epiR package contains the useful epi.interaction() function, which calculates relative excess risk due to interaction, proportion attributable to interaction, and the synergy index (RRAB-1)/(RRAb-1)+(RRaB-1)).
Niklas Luhmann (1927-1998) was a German social scientist working on the "theory of social systems". This centred on decision communications as autopoiesis in organisations, specifically as uncertainty absorption.
Quotes from Luhmann taken from the English language review by Schoeneborn [2011]
Accordingly, social systems are not comprised of persons and actions but of communications...Social systems use communications as their particular mode of autopoietic reproduction. Their elements are communications which are recursively produced and reproduced by a network of communications, and which cannot exist outside the network...Organised social systems can be understood as systems made up of decisions [...]. Decision is not understood as a psychological mechanism, but as a matter of communication, not as a psychological event in the form of an internally conscious definition of the self, but as a social event. That makes it impossible to state that decisions already taken still have to be communicated. Decisions are communications; something that clearly does not preclude that one can communicate about decisions.
Was not translated into English for many years. David Spivak in his work "Toward a mathematical foundation for autopoiesis" cites Luhmann. In Fong et al [2018], they apply topos theory/categorical logic to mereology of (complex) systems.
Exchangeability is a concept that is closely bound up with ideas of individuality and identity. That is to say that one's head starts hurting once one examines it too closely. In probability and statistics, it is a crucial formalized idea that underlies whole areas like Bayesianism and subjective probability. In that mathematical context, exchangeable events or variables can be permuted without altering the properties that you are interested of the that system. A classic example is the number of heads out a series of coin tosses. You are not interested in what order the successes appear, just the number, so any pair of events could have been swapped around in the sequence without altering the relevant outcome.
In the statistical mechanics of quantum physics (QFT), the exchangeability of fundamental particles is more than just a matter of the interests of the observer - they are an essential property of matter. As you might guess, exchangeability is a type of symmetry relation in physics and mathematics. Failure to correctly model exchangeability in (semi-classical) thermodynamics leads to Gibbs Paradox.
The concept of a sortal in philosophy has a similar flavour, in that there is numerosity, with the things being counted identical to each other in respect of the properties used to define the sortal eg "three brown dogs".
So, can two entities be exchangeable without being identical? That is, is it a weaker concept than being perfectly indiscernably alike? I have recently been enjoying Rodin's Axiomatic Method and Category Theory, where in Chapter 5 he runs over some of the complications of defining and using the identity relation. He comments:
We see that Plato, Frege and Geach propose three different views of identity in mathematics. Plato notes that the sense of the "same" as applied to mathematical objects and to the ideas is different: properly speaking, sameness (identity) applies only to ideas while in mathematics sameness means equality or some other equivalence relation. Although Plato certainly recognizes essential links between mathematical objects and Ideas (recall the ideal numbers) he keeps the two domains apart. Unlike Plato Frege supposes that identity is a purely logical and domain- independent notion, which mathematicians must rely upon in order to talk about the sameness or difference of mathematical objects, or any other kind at all. Geach's [Theory of Relative Identity] has the opposite aim: to provide a logical justification for the way of thinking about the (relativized) notions of sameness and difference which he takes to be usual in mathematical contexts and then extend it to contexts outside mathematics..."Any equivalence relation ... can be used to specify a criterion of relative identity."
Rodin then introduces what he variously describes as a constructivist (in that one specifies a procedure) or substantialist interpretation of the identity x = y: that there is an invertible transformation from x to y and y to x. His discussion now segues into categorical logic and category theory (which some folks might be impressed to see is Hegelian), but I will wander off in a simpler non-mathematical way back to exchangeability of x and y.
I am thinking of x and y each sitting within their relationships with everything around them, and we perform symmetrical tranformations of x to y and y to x. Let's say these transformations are simple translations (movements) in space or time. So I might swap one coin for the other coin in two separate sequences of coin throws. In the traditional statistical setup for subjective probability, I might swap coin tosses from different times in a single sequence of throws. And if I were Black (1952) I might have swapped the two identical spheres that are all a particular universe contains.
So at the practical level, no-one involved may notice there has been an exchange, or there were obvious physical changes in whichever object you were paying attention to, or subtle changes in behaviour detectable by prolonged observation eg the statistical properties of the sequence of coin tosses in a change point model. I might consider how much work is required to detect my substitution in an informational sense. In the Leibnitzian idea of identity of indiscernibles, there is an implication that there is an absolute true state of nature (I won't go off into quantum mechanics again) or the observer doing the discernment can spend an infinite amount of time and energy getting to the bottom of things while taking no time at all (like one of those hypercomputers).
Is there a point to this? I see it as a nice way to think about various hoary thought experiments regarding identity:
So is exchangeability just another term for equivalence in these examples? I find it useful because of its constructivist quality - that is it specifies what action is to be taken to demonstrate that two entities are exchangeable. Time travel might be a little hard to realise - in the statistical settings the permutation is epistemic (and counterfactual) rather than ontological. That the testing of equivalence is via checking both sides of the swap is an interesting feature that again speaks to a pragmatic definition rather than just a purely a priori attempt to specify identity or equality, for certain values of pragmatic. It seems to lend itself to interpretations of these various thought experiments that strike me as common-sensical. Where it can't actually answer the question that two entities are actually exchangeable, it can suggest forms of test one might use in the future.
I should ask whether identity has somehow been smuggled in as the identity-preserving transformation. In the comments, we discussed the lump of clay <-> statue. It seems to me that the necessary transformations are quite information heavy in only one direction, that is clay->statue contains all you need to convert any lump of clay to the specified statue, while statue->clay is also suitably general but lots easier to implement. Rodin presents the argument that it is morphisms that are the "real" objects.
Gonzalez-Fortes et al [2017] report results for six Eneolithic and Mesolithic samples (4 Romanian 5.4-8.8 kya, 2 Spanish)
Gomes et al [2017 describe one sample from the Chalcolithic
| ID | Origin | rs12203592 |
|---|---|---|
| SC1_Meso | Rom | C/C |
| SC2_Meso | Rom 8800 bp | T/T |
| OC1_Meso | Rom | C/T |
| GB1_Eneo | Rom | C/T |
| Chan_Meso | Spain | C/C |
| Canes | Spain | C/T |
| Asturias | Spain 3480 bp | C/C |
The rs12203592*T variant is especially common in modern Irish, and is associated with dark hair, light irides and skin, and increased freckling. T carriers are less likely to be wild type for MC1R, at least in the Australian population.
Pirinen M, Donnelly P and Spencer CCA (2012): Including known covariates can reduce power to detect genetic effects in case-control studies. Nat Genet 44: 848-851.
O'Connor and Price [2017, 2018] take advantage of the fact that "if trait 1 is partially genetically causal for trait 2, then most SNPs affecting trait 1 will have proportional effects on trait 2, but not vice versa...".
| SNP effect ak = qkp + gk |
| E(a13a2) =
κpq13 q2 + 3 rg where κp is the kurtosis of p | |
| gcp = log(abs(q2)-log(abs(q1))/
(log(abs(q2)+log(abs(q1))) where gpc=1 when "trait 1 is fully genetically causal for trait 2: q1 = 1 and q2=rg" |
They estimate rg via a LD score bivariate regression, and normalise the estimates of a via a separate LD score regression (jackknifing for standard errors) since:
[o]ur method exploits this excess kurtosis; when κp is zero (such as when p is normally distributed), we are unable to test for partial causality or to estimate gcp (indeed, the model is not identifiable when p is normally distributed...However, it turns out that the Gaussian case is the only non-identifiable case, assuming that (p, g1 , g2) are independent. The following proposition asserts that under an independence assumption, the model is identifiable if and only if does not follow a normal distribution. It does not matter what the marginal distributions of g1 and g2 are. This result echoes similar results in Independent Components Analysis, which separates independent, additive signals exploiting non-Gaussianity....[T]he LCV model includes only a single intermediary and can be confounded in the presence of multiple intermediaries, in particular when the intermediaries have differential polygenicity. Indeed, some trait pairs [in UKBB] appear to show evidence for multiple intermediaries. Nonetheless, causality or partially causality provide a more parsimonious explanation for estimated causal effects, especially when the gcp estimate is high.
Poet (painter, musician, official), lived Tang Dynasty 699-759.
Wingceltis goldenrain shine empty bend
Fresh and green ripple ripples ripples
Secret enter Shang hill road
Woodcutter not able know
Meta RL goes by many different names: learning to learn, multi-task learning, lifelong learning, transfer learning. The goal, however, is usually the same - we wish to train the agents to learn transferable knowledge that helps it generalize to new situations...This problem definition induces an interesting consequence: during meta-learning, we are no longer under the obligation to optimize for maximal reward during training. Instead, we can optimize for a sampling process that maximally informs the meta-learner how it should adapt to new environments. In the context of gradient based algorithms, this means that one principled approach for learning an optimal sampling strategy is to differentiate the meta RL agent's per-task sampling process with respect to the goal of maximizing the reward attained by the agent post-adaption.
Formalism of a reinforcement learner is a Markov Decision Process ({States}, {Actions}, p, γ), with p the dynamics, with p(r,s' | s,a) the probability of reward r, state s' following action a in state s, and γ the discounting of future rewards. The classic architecture of a reinforcement learner is as a LSTM (long short term memory) recurrent layer NN [Hochreiter and Schmidhuber 1997].
AlphaZero is a deep reinforcement learner. "While the results are impressive, they were achieved on one task at the time, each task requiring to train a new agent instance from scratch...Parallel multi-task learning has recently achieved remarkable success in enabling a single system to learn a large number of diverse tasks...We... automatically adapt the contribution of each task to the agent's updates, so that all tasks have a similar impact on the learning dynamics. This resulted in state of the art performance on learning to play all games in a set of 57 diverse Atari games. Excitingly, our method learned a single trained policy - with a single set of weights - that exceeds median human performance. To our knowledge, this was the first time a single agent surpassed human-level performance on this multi-task domain." [Hessel et al 2018]
Wang et al [2018] (Deep Mind) describe a meta-reinforcement-learning algorithm that mimics the pattern of learning of monkeys in the Harlow task, where two stimuli (objects) are presented, one of which hides a reward (in a hole under it). The pair of objects changes in blocks of six, but the rules remain the same. Monkeys learn the abstract rules of the game, so they quickly find the reward regardless of the actual objects. Wang's model is that dopamine represents the local NN reward prediction error (RPE) signal for synaptic learning in eg dorsal striatum thalamus. In the prefrontal cortex a separate network uses the same mechanism, with dopamine (RPE) representing "the expected values of actions, objects and states [by] dynamically [encoding] conversions from reward and choice history to object value, and from object value to object choice". This modulates the former RL process via the corticostriatal pathways - "learning to learn".
Crucially, the emergent algorithm is a full-fledged RL procedure: It copes with the exploration-exploitation tradeoff, maintains a representation of the value function, and progressively adjusts the action policy. In view of this point, and in recognition of some precursor research [Schmidhuber et al, 1996, Simple principles of metalearning], we refer to the overall effect as meta-reinforcement learning.
Added 20171207: New Science paper on more AlphaZero chess games versus Stockfish "A general reinforcement learning algorithm that masters chess, shogi, and Go through self-play".
http://cdclv.unlv.edu/pragmatism/
And Parthmore [2013] on the relationship between Gardenfor's conceptual space and semiotics.
And Hermann-Pillath [2010] on entropy and naturalising semiotics.
Rorty [1989] Contingency, Irony and Solidarity mentions "humour" exactly three times, discussing Nietzsche and Heidegger (!?), and "jokes" (5 times) only when discussing Derrida (and Nabokov).
He takes his epigraph from Kundera's The Art of the Novel:
The agélastes [Rabelais's word for those who do not laugh], the non- thought of received ideas, and kitsch are one and the same, the three- headed enemy of the art born as the echo of God's laughter, the art that created the fascinating imaginative realm where no one owns the truth and everyone has the right to be understood.
Rorty,
...sketches a figure whom I call the "liberal ironist." I borrow my definition of "liberal" from Judith Shklar, who says that liberals are the people who think that cruelty is the worst thing we do. I use "ironist" to name the sort of person who faces up to the contingency of his or her own most central beliefs and desires - someone sufficiently historicist and nominalist to have abandoned the idea that those central beliefs and desires refer back to something beyond the reach of time and chance. Liberal ironists are people who include among these ungroundable desires their own hope that suffering will be diminished, that the humiliation of human beings by other human beings may cease.[...] One of my aims in this book is to suggest the possibility of a liberal utopia: one in which ironism, in the relevant sense, is universal... Solidarity is not discovered by reflection but created. It is created by increasing our sensitivity to the particular details of the pain and humiliation of other, unfamiliar sorts of people.
From Lototsky's notes:
Longest increasing subsequences: Let Z = (Z1..Zn) be i.i.d. U(0,1) random variables and let Sn be the length of the longest increasing subsequence of Z . Then(Sn-2n1/2) / (n1/6) -> TW (as n -> inf)
where TW is the Tracy-Widom distribution. [...]
Largest principal component: The largest principal component of the covariance matrix XT X, where X is an n × p data matrix all of whose entries are independent standard normal has TW as the limiting distribution after proper centering and scaling.
Tracy CA and Widom H. (1994). Level spacing distributions and the bessel kernel. Commun Math Phys. 161 :289--309.
Tracy, C. A. and Widom, H., Distribution Functions for Largest Eigenvalues and Their Applications , ICM 2012, Vol. I, 587-596, 2012.
Distribution functions well approximated by Gamma [ Chiani, 2012].
A common usage is to select the number of significant eigenvalues (smartpca uses it):
Patterson N, Price AL and Reich D. (2006). Population structure and eigenanalysis. PLoS Genet. 2 :20.
And see the review of ancestry programs by Liu et al [2013].
For a general discussion of random matrix theory (from the Wikipedia article), see Deift [2005]. Deift was a coauthor on the longest increasing subsequence paper.
Majumbar and Schehr [2014] discuss this as another extreme value distribution for strongly correlated random variables, specifically
the probability distribution of the top eigenvalue lambda-max of an N × N Gaussian random matrix for large N. While the typical small fluctuations of lambda-max of order ~O(N-2/3) around its mean lambda-max=sqrt(2) are described by Tracy-Widom distributions, atypically large fluctuations of O(1) are described by two different large deviation functions respectively on the left and on the right of the mean. These two tails correspond to very different physics in terms of [an] underlying Coulomb gas describing the eigenvalues: the left large deviation corresponds to a pushed Coulomb gas, while the right large deviation corresponds to a pulled Coulomb gas.
The T-W distribution in this setup represents the distribution around a phase transition of the system from strongly coupled ("unstable") to weakly coupled ("stable").
Very little in the Anglophone literature, but Deleuze and Weber both quote his work. Very similar flavour to the enactivists, though theistic neofinalist existentialist. From Wiklund [1960]:
Ruyer foursquarely maintains that meaning and direction (sens) are inherent in the inorganic and organic world, and that the activity of conscious man, firmly rooted in that world, is its natural prolongation. ...[A]n action, as it unfolds itself in a spatiotemporal world of cause and effect, cannot be understood without reference to its goal. This finality of the action gives meaning to all that is purely a succession of causes and effects in it.
His axiological cogito is "I am seeking a first truth, a rock-bottom certitude, therefore I am free. Freedom is then the certitude I sought, because the search for knowledge implies freedom, which is the positive condition of such a search." [Ruyer, Neo-finalisme 1952].
To deny freedom, pretending in so doing, to enounce a verity, that is ultimately to affirm freedom. The determinist is evidently not pushed into activity a tergo, as he would maintain, but rather initiates an argument to a purpose, and, in the very act, frees himself of the fetters he purports to prove are there. His activity is manifestly not purely mechanistic, but essentially axiological, i.e., it affirms a value; it strives toward a goal.
Neofinalist concepts are existence, freedom, finality, work, invention, value - that is goal-direction, that is life.
Neo-finalisme has a 2016 English edition and the above seems a good summary of his theses. For example, the form of his axiological cogito is based on the double dilemma type argument of Renouvier eg
- Being a pure set of processes, I affirm that my activity is senseless.
- Pursuing senseful ends, I affirm the absurd nature of my activity.
- Being a pure set of processes, I affirm that my activity has a sense.
- Pursuing senseful ends, I affirm that my activity has a sense.
Assertions 1 and 3 eliminate themselves. The fact that assertion 2 is an assertion completely undermines it. So assertion 4 remains.
Translating sens as sense might not be optimal here, given it is also direction and meaning.
From The Liberalism of Fear:
To step outside these customs is not, as the relativist claims, particularly insolent and intrusive. Only the challenge from nowhere and the claims of universal humanity and rational argument cast in general terms can be put to the test of general scrutiny and public criticism.
Left Horse Float car park ~1020
White Rock ~1100
Spring Mountain summit ~1300
Left summit ~1345
Back at car ~1700
Total ~16.3 km.
Didn't rain, though cloudy and hazy to boot. Lots of families on the White Rock summit. We continued along main track to Spring Mountain, which we ascended from the N side.
Excellent views of Brisbane, Ipswich, Springfield. Has two solar-powered transmitters/repeaters and old survey marker. Returned via "Daisy's track", along ridge heading NNW, with little cliffs along and one razorback section. Return along Yuddamun track very undulating - all took longer than expected.
In blossom: banksias, some gums. Lots of smaller birds.
La guerre et le romantisme, fléaux effroyables
War and Romanticism, what terrible scourges!
a Sverdrup (Sv) is a flow of a million cubic meters per second.
If you doubt that the AMOC has weakened, read this
For 1994-2013, Rossby et al. (2013) - at the Oleander line between 32° and 40° North - found a decrease in the upper 2000m transport of the Gulf Stream by 0.8 Sverdrup...
From Ney [2016]:
The preface paradox (Makinson 1965) is that facing the writer of the work of nonfiction who after carefully researching all of the very many claims made in her book recognizes that given human fallibility, it is likely she made at least one mistake somewhere. So to be honest, she acknowledges this fact in her book's preface. In this (common) situation, the author has excellent reason for making each claim in the main text of the book. But she also has excellent reason to believe at least one of these claims is false.
In other words, FDR.
Commodity fetishism is the tendency, in a capitalist commodity system, for social relations between people to appear as a relationship between things (Marx, 1867/1976, p. 164). For example, capital appears to have a life of its own, capable of commanding a return and hiring and firing labor. However, capital is no more than the product of labor, organized and exploited through a social relation between the capitalist and the worker. All commodities appear to trade in relation with each other, to have an inherent value compared to all other commodities, but this merely masks the social relations hidden in the production of these commodities. Therefore, the thing becomes the bearer of value, not the labor that went into its creation (Geras, 1986, p. 59). This makes it appear as though there exists an invisible hand of the market that operates according to scientific laws, outside the realm of human control, when this market is merely the product of human relationships (Sweezy, 1942, p. 36). The organization of production through the purchase of commodities provides a highly effective mask over the exploitative class rela- tionships within a capitalist economy. It appears as though all actors in the economic system are the owners of a commodity, whether a specific product or an input into production, and as such, they stand as equals, each with something to sell in a voluntary exchange (Sweezy, 1942, p. 39).
Hamlyn comments:
...the fundamental mereological principle that the parts that enable a decision maker to make decisions cannot themselves be decision makers.
and Figdor ripostes
Furthermore, and I also argue this at length in the book, the so-called mereological fallacy isn't a fallacy - I suspect it arises by taking compositional principles that appear to hold for objects and extending them, mistakenly, to activities. Elliot Sober also pointed this out long ago. Planets and atomic nuclei both rotate (to use his example, though the point is quite general) - a planet doesn't (and, maybe, cannot) have planets as parts, but there's nothing wrong with a planet rotating and its parts rotating.
Pritchard and Przeworski showed that the noncentrality parameter for the binary trait association Z-score is approximately rλN1/2. Han and Eskin [2011] invert this to estimate the apparent relative risk for the tag SNP (of a similar allele frequency) as:
RRM ~ [((RR-1)p - RR + 1)r + (1-RR)p-1] / [(RR-1)pr+(1-RR)p-1]
They point out that the fixed-effects meta-analysis combining Z values needs to use the SNP allele frequencies to correctly match the results from the full data Mantel-Haenszel test or inverse-variance weighted odds ratio.
Their main proposal is for a new random-effects approach that assumes there is no heterogeneity under the null hypothesis, that is that μ=0 and τ=0. In the presence of heterogeneity, they suggest a genomic control approach where they separately adjust the heterogeneity component and the main effect component of their statistic for multiple testing.
Their Metasoft program is at:
http://genetics.cs.ucla.edu/meta/
Miosge et al [2015] assessed 33 de novo mouse mutations in "essential immune system genes" in vivo and all possible TP53 mutations in vitro measured as p21WAF1 transcriptional enhancer activity. For the mouse mutations:
| PolyPhen | SIFT | ||||
| pheno | Benign | Poss.Dam. | Prob.Dam. | Tolerated | Delet |
| No | 11 (42%) | 8 (31%) | 7 (27%) | 15 (58%) | 11 (42%) |
| Yes | 0 (0%) | 1 (25%) | 3 (75%) | 0 (0%) | 4 (100%) |
These are PolyPhen AUC c=0.81, and SIFT c=0.82.
For TP53:
Of the 1102 mutations predicted to be deleterious with [Polyphen2] score of 0.8 or greater, 42% had good TA activity measured in yeast with a reporter carrying the TP53-binding sequence from p21WAF1. When these 1102 predicted deleterious mutations were tested for activity against TP53-binding sequences from other target genes, the fraction that were FP predictions ranged from 34% for MDM2 sequences to 61% for P53R2 sequences. By contrast, false-negative (FN) predictions, where mutants have less than 50% of WT TA activity yet are predicted not to be damaging (PolyPhen2 0.2), accounted for only 93 of the 2026 mutations (4.6%).
Fire breaks all signposted now. Lots of cobbler's pegs growing beside and often in the middle of roads. Round trip ~11 km (1140-1600) - 2.2 km along Lepidozamia Road to turnoff. Waterfall flowing. Newish shelter shed, water tanks and fireplace at Lepidozamia Park.
Pozdniakova and Ladilov [2018] review the role of soluble adenylate cyclase. The first 9 ADCY genes code the familiar transmembrane adenylate cyclases, whose produced cAMP are found only close to the plasma membrane (phosphodiesterases etc mop up). The cAMP present in organelles and the nucleus are produced by soluble adenylate cyclase. This is sensitive to bicarbonate, so among other roles acts as a pH sensor. Rahman et al [2016] showed that sAC is "essential for lysosomal acidification. In the absence of sAC, V-ATPase does not properly localize to lysosomes, [and] lysosomes fail to fully acidify". Specifically, sAC complexes with V-ATPase in many cell types, and is required for correct PKA-dependent translocation.
"To each is given a bag of tools,
A shapeless mass and a book of rules.
And each must make, ere life is flown,
A stumbling-block or a stepping-stone." R. L. Sharpe,
...we applied whole-exome sequencing to three families with sound-color (auditory-visual) synesthesia affecting multiple relatives across three or more generations. We identified rare genetic variants that fully cosegregate with synesthesia in each family, uncovering 37 genes of interest. Consistent with reports indicating genetic heterogeneity, no variants were shared across families. Gene ontology analyses highlighted six genes - COL4A1, ITGA2, MYO10, ROBO3, SLC9A6, and SLIT2 - associated with axonogenesis and expressed during early childhood when synesthetic associations are formed.
Recently, Johnson [2013] proposed a new method for specifying alternative hypotheses. When used to test simple null hypotheses in common testing scenarios, this method produces default Bayesian procedures that are uniformly most powerful in the sense that they maximize the probability that the Bayes factor in favor of the alternative hypothesis exceeds a specified threshold. A critical feature of these Bayesian tests is that their rejection regions can be matched exactly to the rejection regions of classical hypothesis tests.
Andrew Gelman and many others commented unfavourably at the time.
Test-retest correlation:
| Study | Instrument | r | N | Notes |
|---|---|---|---|---|
| Catron and Thompson [1979] | WAIS | 0.74-0.90 | 76 | 1 wk to 4 mo |
| Snow et al [1988] | WAIS | 0.90 | 101 | 1 y |
| Spitz et al [1983] | WAIS | 0.75 | 42 | Young MR |
| Spitz et al [1983] | WISC | 0.84 | 24 | Young MR |
| Rae and Olson [2018] | IAT | 0.34-0.48 | 519 | Children |
| Bar-Anan and Nosek [2014] | IAT | 0.45 | 116 | 1 h |
Practice effects for the WAIS are well-known, 5-10 points for FSIQ after 3 weeks to 6 month intervals. Estevis et al [2012] suggest that the individual FSIQ level predictive half-interval is 11.
Rae and Olson [2018] comment "[t]he Implicit Association Test (IAT) is increasingly used in developmental research despite minimal evidence of whether children's IAT scores are reliable across time or predictive of behavior".
The IAT-anxiety test has test-retest of 0.5 over 1 year [Egloff et al 2005].
A broad consensus is now emerging that "commensal" and "mutualistic" processes can lead to domestication, whereby both the domesticator and domesticated species seek out and benefit from cohabitation...Many of the species that have ultimately come to inhabit domestic niches are widely considered to have done so largely autonomously; in other words, to have self-domesticated...[phenotypes include] depigmentation; floppy, reduced ears; shorter muzzles; curly tails; smaller teeth; smaller cranial capacities (and concomitant brain size reduction); paedomorphosis; neotenous (juvenile) behavior; reduction of sexual dimorphism (feminization); docility; and more frequent estrous cycles.
A subset of the genes flagged as selected overlapping between humans and other domesticated species:
...BRAF, CACNA1D, NCOA6, LYST, TAS2R16, TP53B1
Minsky in Semantic information processing [1968]:
the science of making machines do the things that would require intelligence if done by men.
Digging a hole? Came up when chasing the Emotions-Beliefs-Desires-Intentions model for agent-oriented programs and agent-based (social) simulations.
Richard Kyle coined the phrase "graphic novel" in CAPA-alpha, an American comic fanzine, in November 1964, and used it to describe "long-form" comic book stories of broadly serious artistic merit. The term was adopted to promote a cluster of science-fiction/fantasy titles, including Richard Corben's Bloodstar (Corben 1976), George Metzger's Beyond Time and Again (Metzger 1976) and Don McGregor and Paul Gulacy's Sabre (McGregor and Gulacy 1978), which were published as paperback books, and distributed via specialty comic shops. However, it was left to Will Eisner's A Contract with God (Eisner 1978) - a work erroneously cited as the first American graphic novel - to demonstrate the medium's potential......popularized 1985 [onwards]...Batman: The Dark Knight Returns... Watchmen...Maus
Bagati et al [2018] (includes Neil Box) previously reported on a role for FOXQ1, previously implicated in numerous tumour types, as a tumour suppressor gene for melanoma. In the present paper, they show FOXQ1 acting upstream from MITF to modulate melanocyte differentiation.
They find the MITF promotor contains FOXQ1 binding sites, confirmed on ChIP. Forskolin was already known to upregulate MITF levels, and they show that depleting FOXQ1 abrogates this (either directly - in culture, or in a FOXQ1 KO mouse model). FOXQ1 may be the intermediary for CREB1 regulation of MITF.
...a potential functional cooperation between FOXQ1 and CREB1 similar to that between SOX10 and its target gene PAX3. Similar to FOXQ1, SOX10 also directly activates MITF [36] and is also required for full-scale activation of MITF by PAX3.Foxq1-null mice retain baseline pigmentation. This observation is consistent with other studies demonstrating that mice deficient of CREB1 or another MITF activator PGC1-alpha also demonstrate "no coat-color" phenotype, underlining the complexity of basal versus induced pigmentation in mouse skin...
UV -> keratinocytes -> MSH -> CREB1 -> MITF
In human and mouse melanocytes, BRAF*600E decreased FOXQ1 (as did depletion of beta-catenin, the most likely downstream target of V600E). Bumping up FOXQ1 levels in these systems seemed to return the cell phenotype to normal.
Germline coding variants in FOXQ1 are not increased in familial melanoma according to Artomov et al [2017]. The best SNPs from the melanoma and nevus metanalyses are rs35434757 (mela P=0.0008) and rs2317902 (mela P=0.000939); no associations in the UKBB for melanoma or pigmentation phenotype.
Callum Hackett comments:
The upshot of this view (known as "meaning eliminativism" in a branch of pragmatics called "relevance theory"), is that words in contexts of use are excellent for referring to concepts, but words themselves do not achieve this by denoting concepts. Instead, words have non-conceptual potentials for conceptual reference, and humans use their pragmatic capacities to fix conceptual referents in context (it is too much to discuss here, but it is important to note that a non-conceptual potential is something quite different from polysemy, which is just a many-to-one correspondence between a word and possible concepts - the claim is instead that words denote something that is materially different from conceptual concept, though use of a word is perceived to have conceptual reference).
Four contextualist positions from Recanati [2004] summarised by Carsten [2012]:
| quasi-contextualism | word meaning may contribute directly, or be modified by context |
| pragmatic composition | word meaning "overwritten" by context |
| "wrong format" | word meanings must be translated into concepts |
| meaning eliminativism | only specific utterings of words have meaning |
Underlying these,
(a) collections of memory traces or exemplars of previous uses (tokenings) and (b) bundles of contingent encyclopaedic information about the things in the world the word is used to refer to.
So how does analogy-making fit in?
During the past few years, various novel statistical methods have been developed for fine-mapping with the use of summary statistics from genome-wide association studies (GWASs). Although these approaches require information about the linkage disequilibrium (LD) between variants, there has not been a comprehensive evaluation of how estimation of the LD structure from reference genotype panels performs in comparison with that from the original individual-level GWAS data. Using population genotype data from Finland and the UK Biobank, we show here that a reference panel of 1,000 individuals from the target population is adequate for a GWAS cohort of up to 10,000 individuals, whereas smaller panels, such as those from the 1000 Genomes Project, should be avoided. We also show, both theoretically and empirically, that the size of the reference panel needs to scale with the GWAS sample size; this has important consequences for the application of these methods in ongoing GWAS meta-analyses and large biobank studies. We conclude by providing software tools and by recommending practices for sharing LD information to more efficiently exploit summary statistics in genetics research.
| Adapt-Mix |
| DISTMIX |
| FAPI |
| HAPRAP |
| ImpG-Summary |
| Popcorn |
...mental representations represent (1) linguistically (in whatever way language represents, without the intervention of sensory modalities); (2) pictorially (in whatever way pictures represent, via quasi-spatial features and the involvement of sensory modalities); (3) a mixture of (1) and (2); (4) magically (they just represent, with no further questions asked on how they do it).Some cognitive psychologists, like Allan Paivio, have defended a "dual coding" theory of mental representations whereby some represent linguistically, some pictorially. People like Kosslyn and others attach great importance to representations of kind (2); others, like Fodor or Pylyshyn, claim that it all reduces to (1)...
I have an argument by cases. Roughly: if the conceivability of P at issue for Humeans involves representation of kind (1), we can conceive the impossible...
People often call "Meinongians" (from the Austrian philosopher Alexius Meinong) those philosophers who claim that some things do not exist... Meinongians are opposed to Quineans, who, on the contrary, claim that everything exists.
His (along with Rossi and Tagliabue) 2010 book The Mathematics of Models of Reference describes an ambitious analysis of a reversible CA system, where the claim is made that recursive self-reference can be implemented:
Recursive self-reference takes place when a MoR [Model of Reference] not only refers to itself, but is aware of such a self-reference. A system implementing a self-referentially recursive MoR can therefore be aware of what it does via that very MoR.Again, this is not semantic animism. For such expressions as "being aware" can be given a precise mathematical meaning. Much research at iLabs is guided by the persuasion that recursive self-reference is at the basis of what people ordinarily call "consciousness": a key difference between conscious thoughts and any other computational procedure is that our mind, as (self-)conscious, can think about and have a viewpoint on itself (albeit with arguably limited powers to operate on its own source code). If Artificial Intelligence is to be real, this will be achieved by means of recursive self-reference - or this is our bet.
The [Kleene (Strong)] Recursion Theorems, applied to our recursive MsoR, guarantee that we can define partial MsoR which are recursively self-referential, for they include their own code in their recursive definition. These are simply classical fixed-point definitions. Since numeric codes are perceptions taken as inputs by (meta-)models of reference, which can also emulate the thought procedures performed by the encoded MsoR, recursive self-referential MsoR can perceive themselves in a precise mathematical fashion, and represent the computational procedure in which they consist within themselves.
Gorfine et al [2017] estimate the SNP heritability via:
Since the lasso possesses the oracle property, it can be shown (details in S1 and S2 Text) that, as the sample size increases, [this estimate] converges to the true heritability value.
They argue that current approaches will underestimate the heritability especially when the number of causative variants is small.
Seminal concept for predictor selection is the sure screening property.
"given n samples for each of p variables, we will use the term "high dimensional" to mean p = O(nk) for some k > 0, and the term "ultra-high dimensional" to mean log(p) = O(nk) [Reese et al 2018]." Saldana and Feng [2017]:
Fan and Lv (2008) introduced a new framework for variable screening via independent correlation learning that tackles the...challenges of ultrahigh dimensional linear models. Their proposed sure independence screening (SIS) is a two-stage procedure; first filtering out the features that have weak marginal correlation with the response, effectively reducing the dimensionality p to a moderate scale below the sample size n, and then performing variable selection and parameter estimation simultaneously through a lower dimensional penalized least squares method such as SCAD or LASSO. Under certain regularity conditions, Fan and Lv (2008) showed surprisingly that this fast feature selection method has a "sure screening property"; that is, with probability tending to 1, the independence screening technique retains all of the important features in the model.Fan and Lv [2010]
As shown in Fan and Fan (2008), even for the independence classification rule described in Section 4.2, classification using all features can be as bad as a random guess due to noise accumulation in estimating the population centroids in high dimensional feature space. Therefore, variable selection is fundamentally important to high dimensional statistical modeling, including regression and classification.
In the correlation screening setup, they show it is possible to reduce the number of included covariates to a submodel with a very high specifiable probability of including all the important predictors, given that a sparse model is correct. Several screening techniques can be shown to have the sure screening property. Once the set is reduced to a moderate scale, other methods can now be used without computational problems. Using a split-half approach, since the sure screening property is present for each analysis, choosing intersecting predictors reduces the false-positive rate.
Fan J, Lv J. Sure independence screening for ultrahigh dimensional feature space. Journal of the Royal Statistical Society: Series B. 2008;70(5):849¿911.
Fan J, Guo S, Hao N. Variance estimation using refitted cross-validation in ultrahigh dimensional regression. Journal of the Royal Statistical Society: Series B. 2012;74(1):37-65.
The R SIS package provides this approach to multiple types of regreesion. The TSGSIS package is an experimental package for pruning SNP pairwise interactions.
Reese et al [2018] discuss these approaches using an extended Cochrane-Armitage test, and compare this to a distance correlation based method. Again, it uses just the screening statistic for each covariate (SNP) in turn.
Pan et al [2017] discuss the Ball correlation, a nonparametric correlation measure (in Banach space). The R Ball package provides a bcorsis() function for screening, using n/log(n) as the default number of variables to retain (or can use jackknifey n-1).
When one object is partly occluded by another, its occluded parts are perceptually "filled in", that is, the occluded object appears to continue behind its occluder. This process is known as amodal completion... the input to visual search is much more complex than previously assumed ...the entry level for vision (that is, entire objects or individual features) can be quite high in many cases,...In our experiments, participants searched for a notched disk target among complete disks and squares. With unlimited exposure duration, when the notched target disk abuts a square (adjacent condition), search is inefficient because the notched target is rendered similar to the complete distractor disks by amodal completion...
that is, the moral standing of future generations.
library(tables)
#
# Dealing with incomplete data
#
N <- function(x) sum(complete.cases(x))
MEAN <- function(x) mean(x, na.rm=T)
SD <- function(x) sd(x, na.rm=T)
f <- formula(cmm * study ~ (n=1) + Height*(mean + sd))
tabular(f, data=subset(x, complete.cases(Height, study ,sex)))
tabs <- list()
tabs[[1]] <- tabular((Factor(BigHx, "History")+1) ~
((n=1) + Format(digits=3)*Percent()),
data=subset(x, !is.na(BigHx)))
tabs[[2]] <- tabular((Factor(Hair_colour)+1) ~
((n=1) + Format(digits=3)*Percent()),
data=subset(x, !is.na(BigHx)))
tabs[[3]] < tabular((1+Factor(sex)) * (Age+TNC5+reflectance_axilla +
reflectance_outer + reflectance_inner) ~
Factor(CM)*(N+Format(digits=2)*(MEAN+SD)), data=x)
f <- "tables.html"
con <- file(f, "wt")
writeLines(myheader, con)
for(i in seq(1,length(tabs))) {
writeLines("", con)
html(tabs[[i]], con)
}
writeLines("