Re: Deprecated Basis Splines functions

Hello, those functions should still be available in version 2.8, but as you note they are marked for deprecation in future releases. Are you not able to use those functions in 2.8? On 7/5/24 10:39, Andrew Forembski via Users list for GNU Scientific Library (GSL) help wrote: [External email -

Re: how to debug multifit_nlinear

Hello, it looks like a problem with the finite-difference approximation to the Jacobian matrix. Can you provide a MWE we can test? On 6/19/24 03:23, PICCA Frederic-Emmanuel wrote: [External email - use caution] Hello, I am using #include in order to do a least square fitting of this funct

GSL use within NASA

Hello all,   If anyone is a NASA employee using GSL, or if anyone knows of a current NASA mission using GSL, can you please contact me? Thanks, Patrick

GSL 2.8 Released

Dear GSL community,   Version 2.8 of the GNU Scientific Library (GSL) has been released. Thank you to all who helped test the library prior to the release, and thank you to everyone for using the library and giving feedback and reports. The following changes have been added to the library: *

Re: Test release for GSL 2.8

this file. Patrick On 5/14/24 16:42, Patrick Alken wrote: Thank you all again for your reports. I have (I think) fixed all the compiler warnings related to filter/movstat, as well as the fabs issues. I would greatly appreciate it if you could do another round of testing on this file: ftp

Re: Test release for GSL 2.8

On 5/12/24 13:34, Patrick Alken wrote: [External email - use caution] All, thank you for your testing reports. I believe I have addressed all the compiler warnings and other issues raised. I have uploaded a new test release here: ftp://alpha.gnu.org/gnu/gsl/gsl-2.7.100.tar.gz I would

Re: Test release for GSL 2.8

mp; make && make check Thanks, Patrick On 5/10/24 16:04, Patrick Alken wrote: [External email - use caution] Dear all,   It is time to make a new GSL release. I have uploaded a test release to: ftp://alpha.gnu.org/gnu/gsl/gsl-2.7.99.tar.gz ftp://alpha.gnu.org/gnu/gsl/gsl-2

Re: Test release for GSL 2.8

Mark,   Thanks for looking at this. A while ago I began to rewrite the cspline interpolation module, to allow it to work for the N=2 case, and also to make it more cache-efficient when evaluating the spline. Unfortunately I never completely finished it and so 2 days ago I reverted everything

Re: Test release for GSL 2.8

thank you for this, I upgraded autoconf to 2.72, which may fix the issue. I will prepare a new test release soon On 5/11/24 13:05, Dave Allured - NOAA Affiliate via Users list for GNU Scientific Library (GSL) help wrote: [External email - use caution] MacPorts has a patch to 2.7.1 that seems

Re: Test release for GSL 2.8

Hello, your suggestions for sparse matrices are good ones. I added a ticket to the bug tracker to track these - I will work on them when I have time On 5/10/24 20:51, Brijesh Upadhaya wrote: [External email - use caution] Hi, All tests passed for gsl-2.7.1+ in Debian 12 with GCC-12.2.0 and

Test release for GSL 2.8

Dear all,   It is time to make a new GSL release. I have uploaded a test release to: ftp://alpha.gnu.org/gnu/gsl/gsl-2.7.99.tar.gz ftp://alpha.gnu.org/gnu/gsl/gsl-2.7.99.tar.gz.sig All reports are welcome - anyone who can test on various platforms would be appreciated (Linux, BSD, Mac OS, Wind

Re: Nonlinear Least-Squares Fitting

Currently, only unconstrained nonlinear least squares is supported. You will need to look into other libraries for constrained methods On 4/23/24 18:17, djvanpeur...@gmail.com wrote: [External email - use caution] Hello All, I'm a new user and have a general question on the trust-region me

Re: basis splines for interpolation (not fitting)

code for this exists on the git repository, but not yet in an official release. If you can download from the git you can look at the examples On 3/8/24 11:56, Simon Wiesheier wrote: Dear all, can I use basis splines for interpolation and not fitting? Or, say I already computed the control poi

Re: Different eigenvectors obtained from gsl_eigen_hermv compared with scipy.linalg.eigh

Note that eigenvectors are unique only up to a constant factor. Since the output of hermv scales the eigenvectors to unit magnitude, they can be scaled by any complex number of unit magnitude and still be a valid eigenvector. It is likely that the GSL and scipy eigenvectors differ by a constant

Re: question about the trust region in the gsl library.

I'm unclear what you are trying to do. Do you simply want to keep track of the trust radius at each step so you can plot it later? On 5/2/23 07:29, Aleja Carmento wrote: Dear Sir/Madam, I am currently working on a gsl least squares method. In this specific case , i have it set to the leven

Re: Reg :: Non Uniform FFT

Have you tried the NFFT library: https://github.com/NFFT/nfft That library is excellent. Also, I think this application is a bit too specialized (and complicated) to include in GSL, especially when there is a good GPL library available (NFFT) Patrick On 3/27/23 08:31, Sumit Adhikari wrote:

Re: fulfill the (cubic) basis splines the partition of unity at all points in [a,b]?

21 Uhr schrieb Patrick Alken : Its also worth mentioning that there has been a substantial overhaul to the B-splines routines since v2.7. The new routines are on the git repository, along with documentation. If you have the ability to clone the git and build the documen

Re: fulfill the (cubic) basis splines the partition of unity at all points in [a,b]?

Its also worth mentioning that there has been a substantial overhaul to the B-splines routines since v2.7. The new routines are on the git repository, along with documentation. If you have the ability to clone the git and build the documentation from there I highly encourage it. There are many

Re: Sparse submatrix view

Hello,   For dense matrices, submatrix views work because all the matrix elements are stored in a nice consecutive, linear manner. This is not the case for sparse matrices. GSL additionally supports three different storage formats for sparse matrices (COO, CSR, CSC). So it is not trivial to i

Re: First ever GSL Technical Report (ALFs)

I like the idea of a DOI. Also the gnu.org website is pretty stable and for citing purposes, I think using the technical report aspect of bibtex with a link to the gnu.org site should be fine. On 2/17/22 11:33, Mark Galassi wrote: I suggest archiving on Zenodo https://zenodo.org/ Zenodo works

Re: First ever GSL Technical Report (ALFs)

Thanks Mark,   I don't think it is suitable for publication in a journal. There is nothing novel here, its just technical details of how to calculate ALFs efficiently. I'll see if I can fix the formatting issue on eq. 36. Its a good idea to cite GSL, I will add that :) Patrick On 2/16/22

First ever GSL Technical Report (ALFs)

Hello all,   I have recently re-written the array versions of the Associated Legendre functions (gsl_sf_legendre) to be much more efficient. The factors in the recurrence relations are now precomputed, and they are stored in a memory layout to maximize cache efficiency. The result is a signif

Re: RANLUX++

Hello,   Yes please do submit a patch. It would be best to submit a git diff against the latest master branch of GSL. I'll take a look when I can find the time. Thanks for your efforts! Patrick On 8/11/21 8:49 AM, Jonas Hahnfeld wrote: Hello GSL developers, over the past months, we (ROOT d

Re: Problem making anonymous contribution

Hello,   I don't know why a git diff would try to push changes upstream. Did you clone an anonymous copy of the gsl git repository? Perhaps you will need to google that error to try to understand what is going wrong. Patrick On 6/12/21 1:32 PM, Carlos Davalillo wrote: Hello I recently got

Re: Copying C array to a GSL vector

First you need to make a "vector view" of your array, and then you can treat it as a vector. Example: gsl_vector_view v = gsl_vector_view_array(my_array, n); gsl_vector_memcpy(dest, &v.vector); On 6/10/21 11:48 PM, Fritz Sonnichsen wrote: I am about to use some filtering routines and I am new

GNU Scientific Library 2.7 released

://ftp.gnu.org/gnu/gsl/gsl-2.7.tar.gz ftp://ftp.gnu.org/gnu/gsl/gsl-2.7.tar.gz.sig The GSL project homepage is http://www.gnu.org/software/gsl/ GSL is free software distributed under the GNU General Public License. Thanks to everyone who reported bugs and contributed improvements. Patrick Alken

Re: Test release for GSL 2.7

========= > > >> Sent: Friday, May 28, 2021 at 9:26 PM >> From: "Patrick Alken" >> To: "help-gsl@gnu.org" >> Subject: Test release for GSL 2.7 >> >> Dear all, >> >>   It is time to make a new GSL

Test release for GSL 2.7

Dear all,   It is time to make a new GSL release. I have uploaded a test release to: ftp://alpha.gnu.org/gnu/gsl/gsl-2.6.90.tar.gz ftp://alpha.gnu.org/gnu/gsl/gsl-2.6.90.tar.gz.sig All reports are welcome - anyone who can test on various platforms would be appreciated (Linux, BSD, Mac OS, Windo

Re: savitzky golay routines?

Currently this is not implemented in GSL On 5/27/21 7:36 PM, Fritz Sonnichsen wrote: > Looking over the routines available from GSL I don't see a lot about > filters. There are some areas of curve fits covered. > I am working with Savitzky-Golay filters at this time--is there no > routines fo

Re: physical constants that might need updating

I logged this email into the bug tracker so it won't be lost. I agree it would be best to stay current with the latest values of the constants. I am concerned that such a change could potentially break existing software which might depend on certain values. Patrick On 5/5/21 11:17 PM, Mark Galass

Re: QR and Pivoted QR for Complex Matrix

Hello,   The complex QR decomposition is available in the git repository - we haven't made a release since it was added. The algorithm is the Level-3 BLAS method of Elmroth and Gustavson, so you need to supply the T matrix (instead of the tau vector) to the routine. If you are willing/able to clo

Re: [PATCH] clarify accelerator usage for multiple interpolations

Thank you, I have added this text to the documentation On 1/29/21 11:26 PM, vishn...@gmail.com wrote: > From: "Vishnu V. Krishnan" > > --- > doc/interp.rst | 10 ++ > 1 file changed, 6 insertions(+), 4 deletions(-) > > diff --git a/doc/interp.rst b/doc/interp.rst > index 5a062d50..9ee31e

Re: Using matrix decompositions

Hello,   There are 2 interfaces for the QR decomposition, the original interface which uses Level-2 BLAS, and a recently added interface using Level-3 BLAS (_r suffix to function names). In both cases, there is an "unpack" function to extract the Q matrix: gsl_linalg_QR_unpack gsl_linalg_QR_unpac

Re: Interpolation accelerator usage

The accelerator works by doing a binary search on the x array, and then caches the index for future lookups. So you can indeed use it for multiple data sets if they share a common x axis. Patrick On 1/28/21 9:54 AM, Vishnu V. Krishnan wrote: > Hi! > > I have a question about allocating accelerato

Re: POTENTIAL SPAM: Speed of complex vs gsl_complex computations

Pavel,   GSL was originally written to conform to the C89 standard, which did not support native complex numbers, so GSL wrote its own implementation of complex numbers and arithmetic. Native support for complex numbers was introduced in C99 and later extended in C11. In the current master branch

Re: eigensystem

gt; eigenvector = >>> -1 +0i >>> 0 +0i >>> 3.00625e-292 +0i >>> >>> >>> which is wrong. >>> The last eigenvector is not correct because this matrix is not >>> diagonalizable. >>> >>> I need to identify such matrices.

Re: eigensystem

disciplinaire Carnot de Bourgogne > 9 Avenue Alain Savary, BP 47870, 21078 DIJON Cedex FRANCE > Tel: +33 (0)380395988 > ======= > > >> Sent: Tuesday, January 19, 2021 at 6:56 PM >> From: "Patrick Alken" >> To: help-gsl@gnu.org >> Subject: Re: eigensys

Re: eigensystem

eigenvalue = 0 +0i >>>>> eigenvector = >>>>> 1 +0i >>>>> 0 +0i >>>>> 0 +0i >>>>> eigenvalue = 0 +0i >>>>> eigenvector = >>>>> 0 +0i >>>>> 1 +0i >>>>> 0 +0i >>>&

Re: eigensystem

What do you mean by handle it? According to the documentation, if the function cannot compute all eigenvalues, an error code is returned. In the case of gsl_eigen_nonsymm, the number of converged eigenvalues is stored in w->n_evals. Patrick On 1/19/21 10:33 AM, Patrick Dupre wrote: > Hello, > > I

Re: GSL Commercial License

Hello,   I don't fully understand whether you want to use GSL in a closed-source app? The answer to that is no, as GSL is released under the GPL. If your app has a GPL-compatible license, then you can use GSL as you wish. Best, Patrick On 1/13/21 1:26 PM, Haris Ali wrote: > Hi, > > I was curiou

New branch for native complex support

Hi all,   FYI there is a new branch on the GSL git repository called 'cplx'. In this branch you will find that you can use gsl_complex as a native C99 complex double type, if you #include prior to including . This is possible because the gsl_complex type is binary compatible with complex double.

Re: How to contribute

Hello,   The easiest way would be to clone the git repository (see GSL webpage), make your additions, and then send me a diff (git diff) file. It would be highly useful to also document your additions (see the doc directory). Thanks, Patrick On 12/14/20 10:34 AM, Tim wrote: > Hello, > I would li

Re: Sum algorithm

Hello,   GSL uses the Welford algorithm, which is based on recurrence relations for mean and variance. See https://en.wikipedia.org/wiki/Algorithms_for_calculating_variance#Welford%27s_online_algorithm Source code here: https://git.savannah.gnu.org/cgit/gsl.git/tree/statistics/mean_source.c htt

Re: Memory limit allocation

Hi Pablo,   The function gsl_spmatrix_alloc needs to allocate space for the non-zero entries of the matrix. By default, it assumes a density of 10%. So if N = 100,000, then it will try to allocate N*N*0.1 = 1e9 elements, each of size sizeof(double), the the total allocation will be 8GB. But then i

Re: Benchmarking GSL

ick On 11/30/20 9:18 AM, Ludovic Courtès wrote: > Hi Patrick, > > Patrick Alken skribis: > >>   Is there a particular part of the GSL library you want to benchmark? >> Some routines have been optimized to run as fast as possible while >> others have not. The overall

Re: Benchmarking GSL

Hello,   Is there a particular part of the GSL library you want to benchmark? Some routines have been optimized to run as fast as possible while others have not. The overall focus of GSL algorithms is on robustness and reliability, not necessarily speed. So you will probably find quite different r

Re: Question on regularized regression.

Hello,   Unfortunately, as you note, support for underdetermined least squares problems is a bit lacking in GSL. GSL's SVD is currently implemented only for M >= N. There may be other ways to solve your problem however. If I understand, you are solving, min_x || b - Ax ||^2 + \lambda^2 || L x ||^

Re: [bug #58977] Any plans to make new release?

Moving this to help-gsl since it is not a bug. There are few patches people have submitted which I need to include. I am also working on updates to the B-spline module which is not quite ready. Its hard to find the time to get everything ready. Is there a particular feature you need which was don

Re: The maximum number of subdivisions was exceeded

I cannot find that error anywhere in romberg.c. It looks more like a QAG error. What type of function are you integrating? Are there singularities? Have you tried increasing n? On 6/19/20 9:35 AM, Patrick Dupre wrote: > Hello, > > Using Romberg, I get: The maximum number of subdivisions was excee

Re: [Help-gsl] Complex QR decomposition

gt; Christian > > > On 11/27/17 10:34 PM, Patrick Alken wrote: >> Hi Christian, >> >>   This should be very useful. I will take a look at your code as soon >> as I can, though it may be a while as I am very busy with other things >> at the moment. I will add this to

Re: [Help-gsl] Complex QR decomposition

clean patches (one > for the Householder fix and one for the complex QR) that could be applied. > > Cheers, > Christian > > > On 11/27/17 10:34 PM, Patrick Alken wrote: >> Hi Christian, >> >>   This should be very useful. I will take a look at your code as soon

Re: Using gsl_matrix_float and gsl_vector_float in for linalg and multiroots functions

Hello,   This is a great idea and should be done at some point, but unfortunately GSL does not currently support this. The gsl_linalg routines are mainly for the double types (gsl_vector and gsl_matrix) and some routines are also supported for complex double (such as Cholesky and LU). If you need

Re: Integration

You could also directly try the QAGI algorithm, though I have never used it: https://www.gnu.org/software/gsl/doc/html/integration.html#qagi-adaptive-integration-on-infinite-intervals On 3/5/20 6:49 AM, Patrick Alken wrote: > Hello, did you try transforming the integral to have finite lim

Re: Integration

Hello, did you try transforming the integral to have finite limits (i.e. https://www.youtube.com/watch?v=fkxAlCfZ67E). Once you have it in this form, I would suggest trying the CQUAD algorithm: https://www.gnu.org/software/gsl/doc/html/integration.html#cquad-doubly-adaptive-integration Patrick O

Re: [Help-gsl] "gsl_spmatrix_float_alloc_nzmax" undefined reference

Your MWE program compiles and runs ok for me. Did you check if you might be linking to an old version of GSL with -lgsl? You may have the new gsl library installed somewhere different. On 10/14/19 4:48 AM, 2592680259 wrote: > I was using "g++ test_gsl.cpp -I/usr/local/include -lgsl -lgslcblas -o

Re: [Help-gsl] Understanding 'ftol' in nonlinear-least-squares fitter

Hello. Only xtol and gtol are currently used by gsl_multifit_nlinear_test(). You can see the code here: http://git.savannah.gnu.org/cgit/gsl.git/tree/multifit_nlinear/convergence.c I made a stub for ftol as well, because the book by Dennis and Schnabel talks about a test for testing the residual

Re: [Help-gsl] gsl-config –version bring the old version

When you run 'make install' in gsl-2.6, it should install gsl-config in your prefix/bin directory, so if you ran: ./configure --prefix=~/usr then gsl-config should be in ~/usr/bin/gsl-config On 8/29/19 12:58 PM, mohamed jabir wrote: > Nothing > > > De : Alan Mead [mailto:am...@alanmead.org] > E

Re: [Help-gsl] latest version of GSL available for Red Hat

GSL can also be easily installed in your home directory, ./configure --prefix=~/usr make && make install Then just make sure you link with the correct library in ~/usr/lib On 8/29/19 10:12 AM, Alan Mead wrote: > You might have better luck asking on a RHEL or CentOS list. In the repos > I have en

[Help-gsl] GNU Scientific Library 2.6 released

://ftp.gnu.org/gnu/gsl/gsl-2.6.tar.gz ftp://ftp.gnu.org/gnu/gsl/gsl-2.6.tar.gz.sig The GSL project homepage is http://www.gnu.org/software/gsl/ GSL is free software distributed under the GNU General Public License. Thanks to everyone who reported bugs and contributed improvements. Patrick Alken

Re: [Help-gsl] test release for GSL 2.6

Hi Sergey,   I don't know what you mean by "polynomial entropy" - can you explain more? Patrick On 8/17/19 11:26 AM, Sergey Shcherbina wrote: > Hi, Patrick, > > Congrats on 2.5.90! > > I have successfully used your GSL for coordinates calculation - the exactness > is equal about > > 0.8-1.2

Re: [Help-gsl] test release for GSL 2.6

Thanks Rhys, I had not thought about other directory builds. I will add that to the todo list for next time :) Patrick On 8/17/19 10:51 AM, Rhys Ulerich wrote: Hey Patrick, Congrats on 2.5.90! ./configure && make && make check Works for me on Ubuntu 18.04 x86_64, additionally trying a VPAT

[Help-gsl] test release for GSL 2.6

Dear all,   I would like to release the next version of GSL (2.6) since there have been significant updates and the usual bug fixes. I have uploaded a test release to: ftp://alpha.gnu.org/gnu/gsl/gsl-2.5.90.tar.gz ftp://alpha.gnu.org/gnu/gsl/gsl-2.5.90.tar.gz.sig All reports are welcome - anyone

Re: [Help-gsl] Interrupting and then resuming minimization

Can you just save the current position vector (state->x) and then start from there the next time? On 8/6/19 5:06 PM, Pedro Gardete wrote: > Hi, > > I'm running a minimization that may need to be interrupted some times, due > to computational constraints. > > Is there a way I can save the "gsl_mult

Re: [Help-gsl] finite element assembly and calling UMFPACK for linear system solution

r now. I need to make this UMFPACK work so I am trying to sort column indices and corresponding values. BR, Brijesh On Wed, Jun 19, 2019 at 10:45 PM Patrick Alken mailto:al...@colorado.edu>> wrote: A new feature on the git is the ability to parallelize the assembly of the sparse matrix. T

Re: [Help-gsl] finite element assembly and calling UMFPACK for linear system solution

d boundary conditions before compressing it. For now I'll just clone and compile the newer version. BR, Brijesh On Wed, Jun 19, 2019 at 7:41 PM Patrick Alken mailto:al...@colorado.edu>> wrote: Hello, On 6/19/19 9:15 AM, Brijesh Upadhaya wrote: > Hi everyone, > > I am working o

Re: [Help-gsl] finite element assembly and calling UMFPACK for linear system solution

Hello, On 6/19/19 9:15 AM, Brijesh Upadhaya wrote: > Hi everyone, > > I am working on a finite element problem and wanted to ask if anyone of you > have tried using UMFPACK for the linear algebra. I have following findings > when using sparse matrix module from gsl-2.5. > > 1. In CCS the row indic

Re: [Help-gsl] permutation matrix with lu decomposition problem

GSL defines the permutation p acting on a vector as (https://www.gnu.org/software/gsl/doc/html/permutation.html#permutations): vi' = P v = v_{pi} So for p = (2,0,1) acting on [ v1 v2 v3]^T, we would obtain [ v3 v1 v2 ] The matrix representation of this is: [ 0 0 1 ] [ 1 0 0 ] [ 0 1 0 ] so it s

Re: [Help-gsl] GNU Scientific Libary

Hello, you can contact me or post to the gsl-discuss mailing list. Patrick On 4/7/19 9:08 PM, Bernard Widynski wrote: > I would like to suggest that a new RNG be included in the GNU Scientific > Library. > > Whom should I contact regarding this? > > Sincererly, > > Bernard Widynski

Re: [Help-gsl] fread error help

I don't see how anyone can help you if you don't provide a minimal working example code On 3/27/19 10:07 AM, Damian Kulec wrote: > Hi, > > I've downloaded the GNU library because i need to do some matrix arithmetic > for a program I am writing. First I need to read in a preexisting binary > file

Re: [Help-gsl] gsl_multifit_nlinear_fdf with combined function and Jacobian?

Olaf,   Ironically, I removed the combined function to make things easier for users. I can confirm that the multifit_nlinear module will always call the Jacobian function after calling the function. So if you want you can store intermediate results in the function call, and then reuse them in

Re: [Help-gsl] non-linear least square fitting

GSL does not currently support constrained optimization directly. You could, however, introduce a large penalty if your parameters are outside the desired range. In your residual function, check if x is outside [x_min,x_max] and set the output f_i to a large value. This will discourage the solver f

Re: [Help-gsl] delete function generates an error in movstat.h

Daniel,   Sorry for the slow reponse. The problem is that "delete" is a C++ keyword which g++ doesn't like in that header file. I always use gcc to compile so I didn't see this problem. Please edit your gsl_movstat.h header file and change "delete" to "delete_sample". Then try compiling your

Re: [Help-gsl] delete function generates an error in movstat.h

Can you provide your source code? On 07/13/2018 07:31 PM, Montezano, Daniel wrote: > Hi, > > I am trying to compile an example from the docs. Compute a rolling mean using > the new package Moving Window Statistics. > > My code is producing the following error (Windows 10, MinGW.org GCC-6.3.0-1,

[Help-gsl] GNU Scientific Library 2.5 released

GNU General Public License. Thanks to everyone who reported bugs and contributed improvements. Patrick Alken * What is new in gsl-2.5: ** doc bug fix in binomial distribution figure (Damien Desfontaines) ** added Wishart distribution (Timothée Flutre) ** added new module

Re: [Help-gsl] test release for GSL 2.5

kk1WehxEWM65f13kMRO48A-bJbqt8gmOYMw07wgccceTCPhW4ueA_CY3-0FGBApMPTrAD4Susd6dG9N845dS9umZ0qxPIymlFSjesbQH_w0faxViTILm347bXYJuBnuDxlfxh4VIvB4JhZL56lBcnnIvQ6AIoChY0tcUpRmlngh_fNCHzaq84hVy4vmseGIWpajhVhAocAVdz0Oc9sBX9Zd3OhH3-n6HZEO4IKQ18pTmhQhIzaPACVOKPKhb7DSe1BCoDFXSbMthY79hiPWvqg2KUFwJNrn1zOmw324MNxj5EbSmI1sbWmtKPabjt1tsEZnbMy4sB859yN7EYgQWJgg0Z0AHx0uUnvw/http%3A%2F%2Fwww.seamplex.com >> >> >> >> On Mon, 20

[Help-gsl] test release for GSL 2.5

Dear all,   I would like to release the next version of GSL (2.5) since there have been some new features added and the usual bug fixes. I have uploaded a test release to: ftp://alpha.gnu.org/gnu/gsl/gsl-2.4.90.tar.gz ftp://alpha.gnu.org/gnu/gsl/gsl-2.4.90.tar.gz.sig All reports are welcome

Re: [Help-gsl] Fwd: Modifying adaptive integration function

8 8:08 AM, Patrick Alken wrote: >>> On 05/10/2018 04:18 AM, Francesco Florian wrote: >>>> Hello! >>>> Since I have received no answer, I wonder whether this is the right >>>> mailing list for this question. If it is not, can you please point me >>&

Re: [Help-gsl] Fwd: Modifying adaptive integration function

On 05/10/2018 04:18 AM, Francesco Florian wrote: Hello! Since I have received no answer, I wonder whether this is the right mailing list for this question. If it is not, can you please point me to the right one? Thank you Hello,   What exactly are you trying to do that QAG cannot? Perhaps the

Re: [Help-gsl] Linear fitting

Francisco,   You have some errors in the way you handle your 2D arrays. I have made a small modification to define new 1D arrays (x and y) and store your fit data into those arrays. The program output now says: Performing fit with 20 points...done! (status = 0) == BET results == y = 0

Re: [Help-gsl] Problems with compiling Non Linear Fit Example

Hello,   I just installed a clean version of GSL v2.4 in my home directory to try your program. Here are the commands I used: > cd gsl-2.4 > ./configure --prefix=/home/palken/usr > make > make install > gcc -g -Wall -o test test.c -lm /home/palken/usr/lib/libgsl.a -lgslcblas > ./test ... summa

Re: [Help-gsl] GSL 2.4 make check failing on Rocks cluster running CentOS 6.9

ithin reported error -3.151853945525253879e-07 2.799408656462224591e-22 FAIL: Bessel Functions [407] On Thu, Jan 18, 2018 at 3:01 PM, Patrick Alken wrote: Hello, It looks like its calculating the value correctly, but the test tolerance needs to be relaxed slightly. Can you locate this l

Re: [Help-gsl] GSL 2.4 make check failing on Rocks cluster running CentOS 6.9

Hello,   It looks like its calculating the value correctly, but the test tolerance needs to be relaxed slightly. Can you locate this line in specfun/test_bessel.c:  186   TEST_SF(s,  gsl_sf_bessel_j2_e, (1048576.0, &r), -3.1518539455252413111e-07, TEST_TOL3, GSL_SUCCESS); and change the TE

Re: [Help-gsl] gsl_integration_qawf

Hello,   A return code of 0 is equivalent to GSL_SUCCESS, so yes that is good. I'm not sure what you mean by inner/outer integration - can you provide a minimal working source code to demonstrate the problem you're having?   The adaptive integration routines in GSL are based on QUADPACK - you can

Re: [Help-gsl] Contribution, PhD thesis topic, Computer Architecture and Scientific Computing

What are your interests? There are plenty of things to work on in GSL, including linear algebra, numerical integration, optimization, RNGs, ODEs, special functions, bug fixing and numerous other topics. On 01/12/2018 09:35 AM, Ali kia wrote: > Hi ! > > I am a first year Computer Architecture PhD s

Re: [Help-gsl] Nonlinear Least-Squares Fitting

Hello,   Yes this is a bug, thanks for reporting it. The problem was I had defined the time variable to be in [0, N], where N is the number of points to fit. Then, for some values of N, after the first iteration the solver finds a negative value for lambda, and so exp(-lambda * N) gets very l

Re: [Help-gsl] Suggested RNG for inclusion in the GNU Scientific Library

Hello,   Do you have an implementation which can interface easily with GSL? If so I can put it on the webpage for people to try out. Patrick On 01/06/2018 08:27 PM, Bernard Widynski wrote: A new RNG based on the middle-square method has been placed in the arXiv. arXiv:1704.00358

Re: [Help-gsl] [Bug-gsl] FFT not generated

oad Srisoonthorn, Thalang, Phuket 83110 Mobil +66(96)926 83 62 Email  peter.scha...@schacht-co.de -Ursprüngliche Nachricht- Von: Patrick Alken [mailto:al...@colorado.edu] Gesendet: Sonntag, 7. Januar 2018 20:37 An: bug-...@gnu.org; peter.scha...@kabelmail.de Betreff: Re: [Bug-gsl] FFT n

Re: [Help-gsl] help in compiling and linking this file

Sorry, the 2.1 docs are texinfo based, so do: cd doc ; make dvi assuming you have texinfo installed On 01/05/2018 02:38 PM, Patrick Alken wrote: > I don't keep them stored online, you can compile them yourself in the > doc/ directory if you have Sphinx installed (try cd doc

Re: [Help-gsl] help in compiling and linking this file

27;t think the integration routines changed between 2.1 and 2.3 On 01/05/2018 02:33 PM, Vasu Jaganath wrote: > Thanks Patrick, where can i read docs for 2.1? Because I think it is > difficult for me to migrate to 2.4 as yet. > > > > On Fri, Jan 5, 2018 at 2:27 PM, Patrick Alken

Re: [Help-gsl] help in compiling and linking this file

That workspace was added in GSL 2.4, so make sure you are using the latest version of the library On 01/05/2018 01:40 PM, Mohammad Akhlaghi wrote: > Hi Vasu, > > The "unknown type name" is not a linker error, its a compiler error (which > comes prior to linking). > > The `gsl_integration_fixed_wo

Re: [Help-gsl] fixed point or adaptive integration for calculating moments using beta PDF?

) behave as x > approaches 0 > or 1? > > On Mon, Jan 1, 2018 at 3:37 AM, Patrick Alken > mailto:al...@colorado.edu>> wrote: > > > The question is whether your Q contains any singularities, > or is highly >

Re: [Help-gsl] fixed point or adaptive integration for calculating moments using beta PDF?

Why not just plot Q(x) over your integration interval? On 12/31/2017 09:00 PM, Vasu Jaganath wrote: > As far as I know, my Q doesn't contain any singularities but I will > recheck my tables again just to confirm. > > Thanks, > Vasu > > On Sun, Dec 31, 2017 at 8:37 PM,

Re: [Help-gsl] fixed point or adaptive integration for calculating moments using beta PDF?

The question is whether your Q contains any singularities, or is highly oscillatory? Is such cases fixed point quadrature generally doesn't do well. If Q varies fairly smoothly over your interval, you should give fixed point quadrature a try and report back if it works well enough for your problem.

Re: [Help-gsl] Complex QR decomposition

Hi Christian,   This should be very useful. I will take a look at your code as soon as I can, though it may be a while as I am very busy with other things at the moment. I will add this to the bug tracker so it isn't lost. Thanks, Patrick On 11/27/2017 08:34 AM, Christian Krueger wrote: Hi,

Re: [Help-gsl] legendre poly deriv

I think the older routine from 1.16 has the same issue. For Legendre derivatives, I believe the current algorithm divides by sin(theta), so at the poles there is a singularity, therefore the routine fails for x=1 or x=-1 as you found. The following paper discusses the problem and proposes an algor

Re: [Help-gsl] Nonlinea fitting with exit info = 27

Hello, there are some tips in the Troubleshooting section of the Nonlinear least squares manual: http://www.gnu.org/software/gsl/doc/html/nls.html#troubleshooting In particular, your problem appears to be an error in your Jacobian matrix. I've attached a modified version of your program where I s

Re: [Help-gsl] Thin QR decomposition

On 10/09/2017 06:03 AM, Ines Arous wrote: > Hi GSL team, > > I am trying to use the QR function that is provided by GSL (which is very > effective compared to other QR implementation I found online). My problem is > that it computes the full decomposition while I need the thin QR > decomposition

Re: [Help-gsl] contribute code for Wishart distribution

Timothée, Yes can you send me a patch against the git master branch? Thanks, Patrick On 07/17/2017 09:45 PM, Timothée Flutre wrote: Hello, I wrote some code to sample random matrices from the Wishart distribution, as well as compute the (log) probability density function, along with tests.

Re: [Help-gsl] test release for GSL 2.4

I tried Mohammad's suggestion, and the tests now write two files in the current working directory: "test.dat" and "test_static.dat". This appears to be working correctly on my system with make check -j8. All function calls to tmpfile() have been removed. It would be great if a windows person can t

Re: [Help-gsl] test release for GSL 2.4

t" and "test-matrix.dat". > > Cheers, > Mohammad > > On June 22, 2017 8:12:06 AM GMT+01:00, Patrick Alken > wrote: > > In GSL 2.3 and earlier, the vector and matrix modules tested the > fwrite/fread routines by creating temporary files with mkst

  1   2   3   >