How can I use Scheme/Lisp/Clojure for Matrix/LP problems?

Question

I need to perform numerical analysis like that supported by MatLab or NumPy. Is there a good library that is supported by Scheme/Lisp/Clojure(Java)? I don't want to leave my round braces.

Thanks a lot.

score 7 · Answer 1 · answered Jul 18 '09 at 08:48

7

For Common Lisp, see

Matlisp: a matrix package for Common Lisp (includes BLAS and LAPACK);
GSLL: GNU Scientific Library for Lisp;

and others on cliki.

answered Jul 18 '09 at 08:48

huaiyuan

26,129
5
57
63

score 6 · Answer 2 · answered Jul 18 '09 at 01:22

6

Incanter is a Clojure-based, R-like statistical computing and graphics environment for the JVM. At the core of Incanter are the Parallel Colt numerics library, a multithreaded version of Colt, and the JFreeChart charting library, as well as several other Java and Clojure libraries.

http://incanter.org/

answered Jul 18 '09 at 01:22

Robert Harvey

178,213
47
333
501

WOW this looks interesting. Are there any benchmarks available? – unj2 Jul 18 '09 at 01:27
+1 for Incanter - been using it for a few weeks with absolutely no issues. Haven't really pushed performance but it seems good. – mikera Feb 22 '11 at 13:01
2

Incanter is using Parallel Colt, which in several benchmarks has shown to be ~10x slower than LAPACK. If anyone has *first-hand experience* with a matrix library for a Lisp using LAPACK/BLAS that has been tested heavily (MATLAB/NumPy/R), please let me know. I have been looking and haven't found it. – gappy Dec 02 '11 at 13:46
Clatrix is a Clojure wrapper of BLAS: worth looking at. It's also a `core.matrix` implementation, which is important as it brings a general purpose N-dimensional array API to Clojure. – mikera Apr 15 '13 at 04:32

score 5 · Answer 3 · edited Jun 20 '20 at 09:12

5

Lisp:

http://www.cs.cmu.edu/afs/cs/project/ai-repository/ai/lang/lisp/code/math/matrix/0.html

Java:

Jama:http://math.nist.gov/javanumerics/jama/
Colt:http://acs.lbl.gov/~hoschek/colt/
Apache commons math: http://commons.apache.org/math/

If speed is important, you can try Parallel Colt: http://sites.google.com/site/piotrwendykier/software/parallelcolt

edited Jun 20 '20 at 09:12

Community

1
1

answered Jul 18 '09 at 01:11

Robert Harvey

178,213
47
333
501

Speed discussion for Java libraries: http://stackoverflow.com/questions/529457/performance-of-java-matrix-math-libraries – Robert Harvey Jul 18 '09 at 01:15
for the lisp library "Last Web update on Mon Feb 13 10:30:23 1995" xD – fortran Jul 30 '09 at 16:42
Math doesn't change much. The Lisp library probably still works fine. – Robert Harvey Jul 30 '09 at 16:51

score 3 · Answer 4 · answered Jul 18 '09 at 02:13

3

For PLT Scheme see the plt-linalg package on planet. There is also mzgsl which has bindings for the GNU Scientific Library.

answered Jul 18 '09 at 02:13

Eli Barzilay

29,301
3
67
110

score 3 · Answer 5 · answered Apr 15 '13 at 04:30

core.matrix is now under development which will bring comprehensive N-dimensional array features (NumPy style) to Clojure: https://github.com/mikera/matrix-api

An important point about core.matrix is that it supports multiple different back-end matrix implementations through the same API.

For example, there is already a pretty good / fast pure Java core.matrix implementation called vectorz-clj, and another one called Clatrix that uses the native JBLAS libraries.

Disclaimer: I'm a major contributor to both core.matrix and vectorz-clj.:

How can I use Scheme/Lisp/Clojure for Matrix/LP problems?

5 Answers5

Lisp:

Java: