package gsl

  1. Overview
  2. Docs
GSL - Bindings to the GNU Scientific Library


Dune Dependency






gsl-ocaml interfaces the GSL (GNU Scientific Library), providing many of the most frequently used functions for scientific computation including algorithms for optimization, differential equations, statistics, random number generation, linear algebra, etc.

Published: 06 Aug 2020


GSL-OCaml - GSL-Bindings for OCaml

GSL-OCaml is an interface to the GSL (GNU scientific library) for the OCaml-language. The currently latest GSL-version known to be compatible is 2.0.


The platform must not align doubles on double-word addresses, i.e. the C-macro ARCH_ALIGN_DOUBLE must be undefined in the OCaml C-configuration header in <caml/config.h>.

Configuring alternative BLAS-libraries

The underlying GSL-library depends on a C-implementation of the BLAS-library (Basic Linear Algebra Subroutines). It comes with its own implementation, gslcblas, which GSL will link with by default.

This implementation is usually considerably slower than alternatives like OpenBLAS or ATLAS (Automatically Tuned Linear Algebra Software) or miscellaneous platform-specific vendor implementations.

If you want GSL-OCaml to link with another BLAS-implementation by default, you will need to set an environment variable before starting the build process (e.g. before opam install):

$ export GSL_CBLAS_LIB=-lopenblas

Note that on Mac OS X GSL-OCaml requires the Apple-specific, highly optimized vendor library vecLib, which is part of the Accelerate-framework, and will automatically link with it. If you do not wish to use Accelerate you can override it; for a Homebrew-installed OpenBlas in the usual place you then need to

export GSL_CBLAS_LIB="-L/usr/local/opt/openblas/lib/ -lopenblas"


Read the GSL manual to learn more about the GNU Scientific Library, and also the GSL-OCaml API.

Usage Hints

Vectors and Matrices

There are several data types for handling vectors and matrices.

  • Modules Gsl.Vector, Gsl.Vector.Single, Gsl.Vector_complex, Gsl.Vector_complex.Single, and the corresponding matrix modules use bigarrays with single or double precision and real or complex values.

  • Modules Gsl.Vector_flat, Gsl.Vector_complex_flat, and the corresponding matrix modules use a record wrapping a regular OCaml float array. This is the equivalent of the gsl_vector and gsl_matrix structs in GSL.

  • Module Gsl.Vectmat defines a sum type with polymorphic variants that regroups these two representations. For instance:

    Gsl.Vectmat.v_add (`V v1) (`VF v2)

    adds a vector in an OCaml array to a bigarray.

  • Modules Gsl.Blas Gsl.Blas_flat and Gsl.Blas_gen provide a (quite incomplete) interface to CBLAS for these types.


Errors in GSL functions are reported as exceptions:

Gsl.Error.Gsl_exn (errno, msg)

You have to call Gsl.Error.init () to initialize error reporting. Otherwise, the default GSL error handler is used and aborts the program, leaving a core dump (not so helpful with OCaml).

If a callback (for minimizers, solvers, etc.) raises an exception, GSL-OCaml either returns GSL_FAILURE or NaN to GSL depending on the type of callback. In either case the original OCaml exception is not propagated. The GSL function will either return normally (but probably with values containing NaNs somewhere) or raise a Gsl_exn exception.

Contact Information and Contributing

Please submit bugs reports, feature requests, contributions and similar to the GitHub issue tracker.

Up-to-date information is available at:

Dependencies (5)

  1. conf-pkg-config build
  2. conf-gsl build
  3. dune-configurator
  4. dune >= "1.10"
  5. ocaml >= "4.08" & < "5.0"

Dev Dependencies


Used by (10)

  1. biotk >= "0.2.0"
  2. curve-sampling >= "0.2"
  3. gpr
  4. libsvm = "0.9.0"
  5. orpie
  6. owl >= "0.2.9" & < "0.3.7"
  7. pareto
  8. phylogenetics
  9. prbnmcn-dagger-gsl < "0.0.3"
  10. prc




Innovation. Community. Security.