Gambit-list Février 2009

gambit-list@iro.umontreal.ca

37 participants
66 discussions

debugging macros
by lowly coder 26 Fév '09

26 Fév '09

(define-macro (define-cool-macro decl . body) `(begin (define-macro ,decl ,@body) (define-macro ,(cons (string->symbol (string-append (symbol->string (car decl)) "-dbg")) (cdr decl)) ,@body))) (define-cool-macro (fn . body) `(lambda ,@body)) (pp (fn-dbg (x) (+ x 2))) Now, does anyone see anything wrong with define-cool-macro, or some way/something that may resulted in unexpected sideeffects (along the lines of accidently introducing variables in macros). Thanks!

1 0

define-type and define-record-type
by Pavel Dudrenov 26 Fév '09

26 Fév '09

I keep seeing references to those in code samples all over gambit written code. What do they do exactly? I think I know, but they seem to take a plethora of keywords as well constructor:, extender:, implementer: etc. Sorry about the noise, Pavel P.S. These are somewhat documented on the wiki, but most of the keywords they take are undocumented.

2 3

Re: [gambit-list] copy-file on windows & links
by Christoph Bauer 26 Fév '09

26 Fév '09

Hi, > I have a question concerning copy-file on windows. > my setup: filesystem on linux mounted over samba on windows I > try to copy a file with "copy-file", which is acutaly a > symlic link under linux. > This results in a .lnk-file. Is there a way to make a real > copy of the file? sorry for the noise, but I think the source of my problem comes from elsewhere. How could I suspect gambit? Christoph Bauer

3 4

optimizing computer simulations of the Ising model
by Marijn Schouten (hkBst) 25 Fév '09

25 Fév '09

Hi Gambitters, I want to use Scheme to implement a computer simulation of the Ising model of spins. Background: http://en.wikipedia.org/wiki/Ising_model Basically all the algorithm does is allocate a vector of spins and flip them randomly in such a way that a representative sampling at a specified temperature is produced. Efficiency of the algorithm is very important as it would allow me to study larger samples. Currently on my own system after compilation: with a sample size of 40x40 I can do 1000 flips/site in a little over 3 seconds. Currently the simulation starts from an all up state and displays some variables and the end configuration of the sample. It looks like this: $ time gsi ising -e "(main 1000)" D: 2, N: 1600, L: 40, BETA*J: .3, RANDOM-STATE: 10624525222961816100342112271285465503733219408383542344109 E: 72, M: -40 +-++++---++---+-+-+-++++--+----+++-+-+-- +++-+---+-+------++---+-+--+--+++++++--- ----+-+--+--+-+--+-+-+++-++---------++-+ ---+---+--++-++++-++++-+--+++-++---+--+- ++++--++-+----++++----+-++---+----+-+++- -----+++--+---+-+++++-++++-+---+++-+++-+ ++-+--+-+---+--+++++-+-+---+++++-++---+- -+---+--++--++--+---++++--+-----+-+-+--+ ----+-+--+---++--+-++-+--+----+++---++-- --++-++---+-++++--+++-++++----+-+++++--- ++-+-+-+-+-+++-+-++++++-+---+---++-+-+-- +--+---++-++++----+-++-+-+-++-++-------+ +--------+++-+-+++---+-+--+-+---++--+-++ +---++-+-+++++++--+-++--+-+++-++++---++- +--+----+++--+--+-+---+++-+-+---+--+++++ ++-+++-+++--+++--+-++++--++-+---+-+-++++ +-+---+++--+-+--+++-+--+---+-++++++-+++- ++---++---+-+--++-+++----+-+++--+++-+--- ++-++-+--+---+++-----+-+----+--+-+-+-+-+ ++--++--+--+-+++--++++--+++++--++-+--+-+ -+---++--+-++--++-+--++-+-++--+-++++-+-+ +----+-+--+++--+---+--+-+++--++--++-+++- ++++--+-++--++++----+-++-+--------+-++++ +-----+++++++-+-+--++-++-+--+++--++-++++ +--++-+-++-------+-+++-----++-+-++-+-++- +--+++--+-+++--+---+---+-----++-+-++-+++ --+-+-+---+-+++----+-+--------+----+-+-+ -+--+--+-+++-+-----+-+----++++-++-++-++- -+-++-+-+--+--+----+++--+-+-+-++--+++-+- +-+++-+-++-+-+++++-+----+-+-+--+-++----+ +------+-+---+-++-+-++++-++-++-++---++-+ ++----++-+-+--+-+-+++-+-+---+++-----++-+ ++++-++++-++-+-+++--++++-++-+--+++++-++- ++-----+++++--++--+-+-+-++-+-+++------+- ++++-+++---+---+-+-++-+---+-+----++++--+ +---++--++---++-+----------+-----+---++- ---++-++--+--++--+--++-+----+--+-++----+ ++--++--+--++++-+---++-++------++-++---- -++-++-++-++---+----+-++---++++++--++-++ +-----+++++--++-+-+---+-++---+-++--++-+- real 0m3.507s user 0m3.198s sys 0m0.007s to compile the program I use: gsc -cc-options "-O2 -march=native" ising I do not know enough about Gambit to know what's fast and what's not and I have no experience with profiling, so I would appreciate some pointers on how to go about optimizing the program. As a side note I was also hoping to compare the performance of various Schemes for this workload, but it seems the ones I tried so far (bigloo and guile) do not support srfi-27 (Sources of Random Bits) much. (It worked in guile until just now when I added a few bits to be able to change the pseudorandom number generator seed.) If anyone knows other implementations where it would run that would be cool. Finally, the source is attached. Any comments appreciated. Thanks in advance, Marijn -- Sarcasm puts the iron in irony, cynicism the steel. Marijn Schouten (hkBst), Gentoo Lisp project, Gentoo ML <http://www.gentoo.org/proj/en/lisp/>, #gentoo-{lisp,ml} on FreeNode ;;;calling ; $ gsc -cc-options "-O2 -march=native" ising ; $ time gsi ising -e "(main 1000)" ;;;end calling ;;;compatibility ;bigloo compat ;(define random-integer random) ;bigloo has no random real generator!!! (define (random-real n) (random ?)) ;guile compat ;guile does not support srfi-27 enough (define random-integer random) (define random-real random:uniform) ;;;end compat (define (SPIN+) +1) (define (SPIN-) -1) ;;; coupling constant (define J 1) ;ferromagnet ;(define J -1) ;antiferromagnet ;;; inverse temperature times coupling constant (define (BETA*J) 0.3) ;(define (BETA*J) 0.4) ;(define (BETA*J) 0.5) (define (display-spin spin) (cond ((= spin (SPIN+)) (display "+")) ((= spin (SPIN-)) (display "-")) (#t display "?"))) ;;; sample length in one dimension (define L 40) ;;; sample dimension (define D 2) ;;; total number of spins ;;; if this is a power of 2 ;;; then modulo can be fast (define N (expt L D)) ;round N down to power of 2 ;(set! N (expt 2 (inexact->exact (floor (/ (log N) (log 2)))))) (define GRID (make-vector N (SPIN+))) (define (display-spatial-sample grid) (let loop ((site 0)) (cond ((< site N) (display-spin (vector-ref grid site)) (if (= (- L 1) (modulo site L)) (newline)) (loop (+ 1 site))))) (if (not (= (expt L D) N)) (newline)) (newline)) (define (display-sample grid) (for-each display-spin (vector->list grid)) (newline)) (define (dimension-loop site kons) (let loop ((d 0) (ret '())) (cond ((< d D) (loop (+ 1 d) (kons d site ret))) (#t ret)))) ;;;helical boundary conditions (define (helical-neighbors site) (dimension-loop site (lambda (d site ret) (cons (modulo (+ (expt L d) site) N) (cons (modulo (- site (expt L d)) N) ret))))) (define (helical-plus-neighbors site) (dimension-loop site (lambda (d site ret) (cons (modulo (+ (expt L d) site) N) ret)))) ;;; Metropolis => equal probabilities for each transition (single-spin-flip) (define (random-site) (random-integer N)) (define (flip! grid site) (vector-set! grid site (* -1 (vector-ref grid site)))) (define (energy-jump/J grid site) (* 2 (vector-ref grid site) (apply + (map (lambda (n) (vector-ref grid n)) (helical-neighbors site))))) ;;; Metropolis => always accept energy-lowering transitions (define (accept-flip grid site) (if (> 0 (energy-jump/J grid site)) #t (< (random-real) (exp (* (BETA*J) (energy-jump/J grid site)))))) (define (single-spin-flip! grid) (let ((site (random-site))) (if (accept-flip grid site) (flip! grid site)))) (define (magnetization grid) (apply + (vector->list grid))) (define (energy/J grid) (define (-energy/J-plus site) (* (vector-ref grid site) (apply + (map (lambda (n) (vector-ref grid n)) (helical-plus-neighbors site))))) (let loop ((s 0) (E 0)) (cond ((< s N) (loop (+ s 1) (- E (-energy/J-plus s)))) (#t E)))) (define (energy grid) (* J (energy/J grid))) (define (time-step) (let loop ((n 0)) (cond ((< n N) (single-spin-flip! GRID) (loop (+ n 1)))))) (define (display-params . p) (let loop ((p p)) (cond ((not (null? p)) (display (car p))(display ": ")(display (cadr p)) (cond ((not (null? (cddr p))) ; not last elt? (display ", ") (loop (cddr p))))))) (newline)) (define (display-config) (display-params "D" D "N" N "L" L "BETA*J" (BETA*J) "RANDOM-STATE" (random-source-state-ref default-random-source))) (define TIMES #f) (define (main args) (random-source-pseudo-randomize! default-random-source 0 0) ; (write args)(newline) (set! TIMES args) (display-config) ; (display-spatial-sample GRID) (let loop ((t 0)) (cond ((< t TIMES) (time-step) (loop (+ t 1))))) (display-params "E" (energy GRID) "M" (magnetization GRID)) (display-spatial-sample GRID) )

2 2

Re: [gambit-list] optimizing computer simulations of the Ising model
by olaf.rogalsky＠web.de 25 Fév '09

25 Fév '09

Depending on what you are trying to do, great speed improvements can be achieved by "entropic sampling"/"importance sampling", where mircocanonical phase states are visited with a propability, which is inverse proportional to the density of states. There is a large literature on this topic, for example: http://en.wikipedia.org/wiki/Importance_sampling http://www.cms.tuwien.ac.at/media/pdf/publications/PRE_71_066705_2005.pdf http://hrma.physics.sjtu.edu.cn/SimulationPhysics/PRL02050.pdf Olaf -- Using Opera's revolutionary e-mail client: http://www.opera.com/mail/

1 0

copy-file on windows & links
by Christoph Bauer 25 Fév '09

25 Fév '09

Hi, I have a question concerning copy-file on windows. my setup: filesystem on linux mounted over samba on windows I try to copy a file with "copy-file", which is acutaly a symlic link under linux. This results in a .lnk-file. Is there a way to make a real copy of the file? Thanks for help, Christoph Bauer

1 0

Difficulties linking against a C++ library
by John Thompson 24 Fév '09

24 Fév '09

Hi, I'm trying to get a toy, loadable C++ library compiled. My setup is as follows: foo.scm: (define bar (c-lambda (char-string) int "cpp_bar")) foo.cpp: int cpp_bar(char* text) { return 0; } foo.h: int cpp_bar(char* text); I'm using the latest tarball of gambit, compiled with --enable-cplusplus. I compile foo.scm to C++ via: gsc -link -flat foo.scm ==> foo.c and foo_.c Then I want to compile these into object files and I run into trouble. g++ -fPIC -c foo.c outputs the following: oo.c: In function ‘int ___H__20_foo_23_0(___processor_state_struct*)’: foo.c:119: error: ‘cpp_bar’ was not declared in this scope foo.c: At global scope: foo.c:156: warning: deprecated conversion from string constant to ‘char*’ Whereas gcc compiles it fine. It must be a name-mangling issue, but it seems peculiar to me, given that I compiled gsc using g++. Moreover, I can compile if I add extern "C" {int cpp_bar(char*);} to foo.c. I picked this up from reading about someone who had similar problems a few years ago: http://osdir.com/ml/lisp.scheme.gambit/2005-02/msg00037.html I then compile and link like: g++ -fPIC -c -D___DYNAMIC foo_.c foo.c g++ -fPIC -c -D___DYNAMIC foo.cpp -o cppfoo.o ld -G -o foo.o1 /usr/lib/libstdc++.so.6.0.9 /usr/lib/gcc/i486-linux-gnu/4.2/crtbegin.o cppfoo.o foo_.o foo.o Unfortunately, this leads me here: *** WARNING -- Could not find C function: "____20_foo_2e_o1" *** ERROR IN ##main -- .../foo.o1: undefined symbol: cpp_bar (load "foo.o1") Can anyone offer me any pointers on how to proceed? -John

4 5

sllgen for gambit?
by Pavel Dudrenov 24 Fév '09

24 Fév '09

Does anyone know if there is anything portable written, or that works on gambit? The only thing I found works is plt scheme. Thanks, Pavel

2 2

Getting rid of compiler warnings.
by lowly coder 24 Fév '09

24 Fév '09

Hi! There are two compiler warnings I want to get rid of. Please enlighten me. (1) is from gcc, (2) is from gsc. (1) this one deals with const char * vs char * if I try to do (pointer char), I get the following error: ~/test$ cat error1.scm; echo '-----'; gsc error1.scm; echo '-----'; gsc -link error1; gsi -e '(load "error1")' (c-declare #<<c-declare-end #include <stdio.h> void foo(const char* str) { fprintf(stderr, "printing: %s\n", str); } c-declare-end ) (define foo (c-lambda ((pointer char)) void "foo")) (foo "hello world") ----- error1.c:45: warning: deprecated conversion from string constant to ?char*? error1.c:51: warning: deprecated conversion from string constant to ?char*? error1.c:51: warning: deprecated conversion from string constant to ?char*? error1.c:51: warning: deprecated conversion from string constant to ?char*? error1.c:191: warning: deprecated conversion from string constant to ?char*? ----- *** ERROR IN ##load -- (Argument 1) Can't convert to C pointer (foo "hello world") now, if I use char-string instead, I get an annoying error while compiling with gcc: ~/test$ cat error2.scm; echo '-----'; gsc error2.scm; echo '-----'; gsc -link error2; gsi -e '(load "error2")' (c-declare #<<c-declare-end #include <stdio.h> void foo(const char* str) { fprintf(stderr, "printing: %s\n", str); } c-declare-end ) (define foo (c-lambda (char-string) void "foo")) (foo "hello world") ----- error2.c:42: warning: deprecated conversion from string constant to ?char*? error2.c:42: warning: deprecated conversion from string constant to ?char*? error2.c:42: warning: deprecated conversion from string constant to ?char*? error2.c:174: warning: deprecated conversion from string constant to ?char*? ----- printing: hello world I don't want to tell gcc to ignore this error. I want to get gambit to generate correct code. How do I do that? (2) This is shown right in section 3.4.2 of the gambit reference manual: $ gsc -link -flat -o foo.o1.c m2 m3 m2: m3: *** WARNING -- "cons" is not defined, *** referenced in: ("m2.c") *** WARNING -- "map" is not defined, *** referenced in: ("m3.c") *** WARNING -- "newline" is not defined, *** referenced in: ("m3.c") *** WARNING -- "write" is not defined, *** referenced in: ("m3.c") Is there something I can put in m2.scm / m3.scm like "don't worry about cons / map / newline / write" ? ; File: "m2.scm" (c-declare "extern int power_of_2 ();") (define pow2 (c-lambda (int) int "power_of_2")) (define (twice x) (cons x x)) ; File: "m3.scm" (write (map twice (map pow2 '(1 2 3 4)))) (newline) aside: I realize these are just small warnings. However, I'm doing alot of C-FFi library wrapping, and it's getting to be more than one screenful of warnings. This is bad as (1) I like things to be clean and (2) it makes it hard to view actual errors since there are so many silly warnings. Thanks!

2 2

how does make-default-entry-hook work?
by lowly coder 23 Fév '09

23 Fév '09

In lib/_repl.scm: (define-prim (##make-default-entry-hook) (let ((settings (##vector #f #f))) (lambda (proc args execute) (if (##vector-ref settings 0) (##step-on)) ;; turn on single-stepping (if (##vector-ref settings 1) (##trace-generate (##make-call-form proc (##argument-list-remove-absent! args '()) ##max-fixnum) execute #f) (execute))))) how does this code work at all? isn't settings #f #f, which would mean the first if does nothing, and the first part of the second if also does nothing? is there some funky dynamic-wind / dynamic scope going on here?

3 3

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

Gambit-list Février 2009