Alexander J. wrote:
I have quite good experience with the bindings to Gnu Scientific Library
(GSL: http://www.gnu.org/software/gsl/manual/html_node/, rb-gsl:
http://rb-gsl.rubyforge.org/). there is also another bindings package:
As you may know, narray (http://narray.rubyforge.org/SPEC.en) is a fast
multidimensional array implementation, and it is interoperable with
rb-gsl (http://rb-gsl.rubyforge.org/narray.html). If you install
‘plotutils’ and ‘gnuplot’ (+ its ruby gem), you will be able to use
GSL’s plotting facilities (http://rb-gsl.rubyforge.org/vector.html#3.17,
graph -> plotutils, plot -> gnuplot). After all, it felt like using
matlab to me. Having a ruby as a fully features language by your side
makes it even more powerfull.
If you’re looking for statistical analysis, rsruby
(http://rubyforge.org/projects/rsruby/), ruby bindings to R.
Another important issue is IO. If you have CSV-like files, GSL can read
the directly, or you can let ruby to that. If you have netcdf
http://ruby.gfd-dennou.org/products/ruby-netcdf/) , grib
(http://raa.ruby-lang.org/project/ruby-grib/) or hdf5
(http://www.hdfgroup.org/HDF5/), you have to look for libraries.