Re: [Help-gsl] test release for gsl-2.0
Both ./configure && make && make check and cd doc; make ps work for me. Tested on Parabola GNU/Linux with gcc 5.2.0 signature.asc Description: PGP signature
Re: [Help-gsl] GSL performance compared to C++ alternatives such as Armadillo
> Armadillo would be faster. > GSL would be more reliable and wider used. Eigen may be another good alternative for C++. - Rhys
Re: [Help-gsl] Linear least squares, webpages and the next release
For everyone who said they’d like to contribute, but feel like their coding isn’t up to GSL standards: don’t worry! This isn’t the Linux kernel, you won’t get called stupid for doing your best. Patrick and the others are generally helpful and will help you to get your code included. I think it would be a shame if potential improvements to GSL go undone because everyone has impostor syndrome. I work almost exclusively in C++ and Python (and I’m a physicist!) and I was able to contribute a new 1D interpolation routine that will be in 2.0. I looked at the existing code as a model and wrote the new routine accordingly. After submitting a patch I had to make some modifications to clean it up, but it was fine! Jean-François On Oct 26, 2015, at 08:13 , Patrick Alken wrote: > On 10/24/2015 09:35 AM, John D Lamb wrote: >> >>> >>> 3. Are you willing to develop and contribute the features you want? >> >> Yes. I’d be happy to look at redesign of the GSL containers. What’s needed? >> > > There was a discussion on gsl-discuss some time back, see: > > https://www.sourceware.org/ml/gsl-discuss/2014-q2/ > > Gerard may have already done some work on this, or have some ideas on a good > starting point, so I suggest getting in touch with him too (cc'd). > > Thanks! > Patrick >
Re: [Help-gsl] GSL performance compared to C++ alternatives such as Armadillo
Armadillo would be faster. GSL would be more reliable and wider used. -Original Message- From: help-gsl-bounces+vsakharuk=hbk@gnu.org [mailto:help-gsl-bounces+vsakharuk=hbk@gnu.org] On Behalf Of Christopher Choi Sent: Monday, October 26, 2015 11:44 AM To: help-gsl@gnu.org Subject: [Help-gsl] GSL performance compared to C++ alternatives such as Armadillo Hello all, First of this is not a flame post, I would like to use GSL as linear algebra library for Neural Network training, so this would involve calculating hessian matrices, normal matrix and vector operations etc... I was curious as to how does GSL compares to C++ alternatives such as Armadillo, what are the pros and cons of using GSL as suppose to the alternatives? Additionally if someone could put me towards some benchmarks that would be great. Thanks Chris
[Help-gsl] GSL performance compared to C++ alternatives such as Armadillo
Hello all, First of this is not a flame post, I would like to use GSL as linear algebra library for Neural Network training, so this would involve calculating hessian matrices, normal matrix and vector operations etc... I was curious as to how does GSL compares to C++ alternatives such as Armadillo, what are the pros and cons of using GSL as suppose to the alternatives? Additionally if someone could put me towards some benchmarks that would be great. Thanks Chris
Re: [Help-gsl] test release for gsl-2.0
On 10/26/2015 09:12 AM, Jorge Fernandez-de-Cossio-Diaz wrote: Where can I see the changes introduced in the new version? Are there new functions supported? http://git.savannah.gnu.org/cgit/gsl.git/tree/NEWS
Re: [Help-gsl] Linear least squares, webpages and the next release
On 10/24/2015 09:35 AM, John D Lamb wrote: 3. Are you willing to develop and contribute the features you want? Yes. I’d be happy to look at redesign of the GSL containers. What’s needed? There was a discussion on gsl-discuss some time back, see: https://www.sourceware.org/ml/gsl-discuss/2014-q2/ Gerard may have already done some work on this, or have some ideas on a good starting point, so I suggest getting in touch with him too (cc'd). Thanks! Patrick
Re: [Help-gsl] test release for gsl-2.0
Where can I see the changes introduced in the new version? Are there new functions supported? On Sun, Oct 25, 2015 at 7:28 PM, Rhys Ulerich wrote: > > Please try testing the build and also the documentation: > > > > 1) ./configure ; make ; make check > > Passes CFLAGS=-O3 ./configure && make -j && make check for gcc (Ubuntu > 4.9.2-10ubuntu13) 4.9.2. > > > 2) cd doc ; make ps > > Works. Also, Patrick, thank you for the kind credit in the manual. I > didn't realize you had added me until my name popped up on the front. > > - Rhys > >
Re: [Help-gsl] GSL 2D interpolation
On 10/26/2015 06:40 AM, Lai Zhong Yuan wrote: Dear all, Recently looking through the GSL user manual i found out that David Zaslavsky's interp2d routine has been included in GSL; however, it seems that it will only be there in "gsl-2.0", according to http://git.savannah.gnu.org/cgit/gsl.git/tree/NEWS (i have v1.15). Does anyone know how i can obtain a working update of gsl that includes "gsl_interp2d.h" and "gsl_spline2d.h"? I tried including these files and other dependencies as best as i could by hand but this yields, as probably expected, errors when linking in GCC (no problems in compiling). Regards,Lai For now you can work from the git: git clone git://git.savannah.gnu.org/gsl.git gsl-2.0 will hopefully be released within the week.
Re: [Help-gsl] Linear least squares, webpages and the next release
Hello List: On 26/10/15 06:27, Foivos Diakogiannis wrote: > b. I know it is too much to ask, but I'd love to see a c++ implementation > of GSL (not just a wrapper), Why C++ ? An effort for an implementation in D would not be better ? Jerome
Re: [Help-gsl] Linear least squares, webpages and the next release
> > In comparison, see http://www.holoborodko.com/pavel/numerical-methods/numerical-integration/ from Pavel Holoborodko for fast GL quadratures. > I'm not an expert on numerical integration so hopefully someone else can come forward and improve on these algorithms :-) Pavel let us add some of his routines awhile back. Look for "glfixed" in the integration documentation. - Rhys
