Hi, Unfortunately I don't think those notes are particularly helpful in this regard: for example, does bundling the software with SciPy meet the requirement of "distributing the ORIGINAL package", or not? It's murky, and most of the devs would likely err on the side of safety and assume the answer is no. The benefit of using an established license like BSD, MIT, GPL, etc. is that the language is well-defined and the intent of the license is well-understood. Jake Jake VanderPlas Senior Data Science Fellow Director of Research in Physical Sciences University of Washington eScience Institute On Fri, Oct 7, 2016 at 11:24 AM, Dieter Werthmüller <dieter@werthmuller.org> wrote:

Jake,

Thanks for your reply.

Have you seen my whole section on "Files and Licences/Permissions"? (I put that info at the end of my original message, after my name.) Do you think it is not enough if Hamilton gave his written permission to 'use fftlog in any way you choose'?

Dieter

...

Hi Dieter, That's very cool! Before discussion of whether this fits into SciPy, there's one issue: I don't see any license on the Fortran code. This means, unfortunately, that it defaults to some form of "all-rights-reserved" and cannot be used in SciPy.

Often it's enough to email the package author, link to some information like my post at [1], and request that they add a BSD-style license to their code (note that a GPL-style license would make it unusable by SciPy).

Though we may want to see what others think about including this in SciPy before going too far down that route. For what it's worth, I'd suggest starting by making sure your Python wrapper is well-documented & well-tested; if it proves useful to many people, it would then be quite easy to pull into SciPy.

Jake

[1] http://www.astrobetter.com/blog/2014/03/10/the-whys-and- hows-of-licensing-scientific-code/

Jake VanderPlas Senior Data Science Fellow Director of Research in Physical Sciences University of Washington eScience Institute

On Thu, Oct 6, 2016 at 7:20 PM, Dieter Werthmüller <dieter@werthmuller.org <mailto:dieter@werthmuller.org>> wrote:

Dear SciPy-devs,

I recently had the need for a logarithmic FFT routine, and did a quick f2py around Andrew Hamilton's FFTLog: http://casa.colorado.edu/~ajsh/FFTLog <http://casa.colorado.edu/~ajsh/FFTLog>

I put my f2py-setup and my pyf-file up on GitHub: https://github.com/prisae/fftlog <https://github.com/prisae/fftlog

I thought it might be a useful addition to the SciPy FFTPack. Specifically as three out of the five fortran-files, of which FFTLog consists, are already in scipy/fftpack.

I have never contributed to SciPy and therefore do not know how much work it would involve to get it into SciPy. However, I thought I ask. If there is interest I expect that it would not take long for an experienced person, as it is a fairly small addition. Or someone could point me to the right direction on what to do to get it in.

I append some comments on the involved files of FFTLog and some comments regarding their licenses. I believe it is compatible with SciPy.

Thanks for all your good work! Dieter

Files and Licenses/Permissions ------------------------------

Files of FFTLog

[1] cdgamma.f [2] drfftb.f [3] drfftf.f [4] drffti.f [5] fftlog.f

(plus a test routine, fftlogtest.f)

-- [1] -- The original FFTLog states about this file:

FFTLog uses [...] and a modified version of the complex Gamma function from the gamerf package at momonga.t.u-tokyo.ac.jp/~ooura/gamerf.html <http://momonga.t.u-tokyo.ac.jp/~ooura/gamerf.html>. The original gamerf copyright statement states:

Copyright(C) 1996 Takuya OOURA (email: ooura@mmm.t.u-tokyo.ac.jp <mailto:ooura@mmm.t.u-tokyo.ac.jp>). You may use, copy, modify this code for any purpose and without fee. You may distribute this ORIGINAL package.

Permission to distribute the modified gamma function code with the FFTLog package has been granted (email from Takuya Ooura to Andrew Hamilton dated 16 March 1999).

Hence I think it is compatible with SciPy, or am I wrong?

Alternatively it could be replaced with scipy.special.loggamma, I think, but I do not know if it is possible to mix Fortran and Python code with f2py.

-- [2], [3], [4] -- They are from the NCAR suite of FFT routines (Swarztrauber 1979).

As far as I can see they are already in SciPy, in scipy/fftpack/src/dfftpack/ (dfftb.f, dfftf.f, and dffti.f).

So there is definitely not a licensing problem here.

-- [5] -- This is the actual FFTLog routine from Hamilton.

I asked Hamilton for permission before publishing his source code at github.com/prisae/fftlog <http://github.com/prisae/fftlog>.

His email response was:

-- start email from 28/09/2016 -- Dieter,

You are welcome to use fftlog in any way you choose. Please note

On 07/10/16 12:53, Jacob Vanderplas wrote: the

...

credits commented in the code:

c FFTLog uses the NCAR suite of FFT routines, c and a modified version of the complex Gamma function c from the gamerf package at c http://momonga.t.u-tokyo.ac.jp/~ooura/gamerf.html <http://momonga.t.u-tokyo.ac.jp/~ooura/gamerf.html> . c The original gamerf copyright statement states: c Copyright(C) 1996 Takuya OOURA (email: ooura@mmm.t.u-tokyo.ac.jp <mailto:ooura@mmm.t.u-tokyo.ac.jp>). c You may use, copy, modify this code for any purpose and c without fee. You may distribute this ORIGINAL package. c c Permission to distribute the modified gamma function code c with the FFTLog package has been granted c (email from Takuya Ooura to Andrew Hamilton dated 16 March 1999).

Andrew -- end email from 28/09/2016 --

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org> https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev>

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org https://mail.scipy.org/mailman/listinfo/scipy-dev

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org https://mail.scipy.org/mailman/listinfo/scipy-dev

<dieter@werthmuller.org> wrote:

Jake,

Thanks for the clarification. I will try to get the permissions from the authors.

What is regarded as sufficient? Is an email from the author, granting me/SciPy to distribute their code with a specific, BSD-style license sufficient? Or do they necessarily have to change the websites where they host the code to include the license?

Best would be a license in the software repository/tarball itself. Jake

Dieter

...

Hi, Unfortunately I don't think those notes are particularly helpful in this regard: for example, does bundling the software with SciPy meet the requirement of "distributing the ORIGINAL package", or not? It's murky, and most of the devs would likely err on the side of safety and assume the answer is no. The benefit of using an established license like BSD, MIT, GPL, etc. is that the language is well-defined and the intent of the license is well-understood. Jake

Jake VanderPlas Senior Data Science Fellow Director of Research in Physical Sciences University of Washington eScience Institute

On Fri, Oct 7, 2016 at 11:24 AM, Dieter Werthmüller <dieter@werthmuller.org <mailto:dieter@werthmuller.org>> wrote:

Jake,

Thanks for your reply.

Have you seen my whole section on "Files and Licences/Permissions"? (I put that info at the end of my original message, after my name.) Do you think it is not enough if Hamilton gave his written permission to 'use fftlog in any way you choose'?

Dieter

On 07/10/16 12:53, Jacob Vanderplas wrote: > Hi Dieter, > That's very cool! Before discussion of whether this fits into SciPy, > there's one issue: I don't see any license on the Fortran code. This > means, unfortunately, that it defaults to some form of > "all-rights-reserved" and cannot be used in SciPy. > > Often it's enough to email the package author, link to some information > like my post at [1], and request that they add a BSD-style license to > their code (note that a GPL-style license would make it unusable by SciPy). > > Though we may want to see what others think about including this in > SciPy before going too far down that route. For what it's worth, I'd > suggest starting by making sure your Python wrapper is well-documented & > well-tested; if it proves useful to many people, it would then be quite > easy to pull into SciPy. > > Jake > > [1] http://www.astrobetter.com/blog/2014/03/10/the-whys-and- hows-of-licensing-scientific-code/ <http://www.astrobetter.com/blog/2014/03/10/the-whys-and- hows-of-licensing-scientific-code/> > > Jake VanderPlas > Senior Data Science Fellow > Director of Research in Physical Sciences > University of Washington eScience Institute > > On Thu, Oct 6, 2016 at 7:20 PM, Dieter Werthmüller > <dieter@werthmuller.org <mailto:dieter@werthmuller.org> <mailto:dieter@werthmuller.org <mailto:dieter@werthmuller.org>>> wrote: > > Dear SciPy-devs, > > I recently had the need for a logarithmic FFT routine, and did a > quick f2py around Andrew Hamilton's FFTLog: > http://casa.colorado.edu/~ajsh/FFTLog <http://casa.colorado.edu/~ajsh/FFTLog> > <http://casa.colorado.edu/~ajsh/FFTLog <http://casa.colorado.edu/~ajsh/FFTLog>> > > I put my f2py-setup and my pyf-file up on GitHub: > https://github.com/prisae/fftlog <https://github.com/prisae/fftlog> <https://github.com/prisae/fftlog <https://github.com/prisae/fftlog>> > > I thought it might be a useful addition to the SciPy FFTPack. > Specifically as three out of the five fortran-files, of which FFTLog > consists, are already in scipy/fftpack. > > I have never contributed to SciPy and therefore do not know how much > work it would involve to get it into SciPy. However, I thought I > ask. If there is interest I expect that it would not take long for > an experienced person, as it is a fairly small addition. Or someone > could point me to the right direction on what to do to get it in. > > > I append some comments on the involved files of FFTLog and some > comments regarding their licenses. I believe it is compatible with > SciPy. > > Thanks for all your good work! > Dieter > > > Files and Licenses/Permissions > ------------------------------ > > Files of FFTLog > > [1] cdgamma.f > [2] drfftb.f > [3] drfftf.f > [4] drffti.f > [5] fftlog.f > > (plus a test routine, fftlogtest.f) > > -- [1] -- > The original FFTLog states about this file: > > FFTLog uses [...] and a modified version of the complex Gamma > function from the gamerf package at > momonga.t.u-tokyo.ac.jp/~ooura/gamerf.html <http://momonga.t.u-tokyo.ac.jp/~ooura/gamerf.html> > <http://momonga.t.u-tokyo.ac.jp/~ooura/gamerf.html <http://momonga.t.u-tokyo.ac.jp/~ooura/gamerf.html>>. The original > gamerf copyright statement states: > > Copyright(C) 1996 Takuya OOURA (email: ooura@mmm.t.u-tokyo.ac.jp <mailto:ooura@mmm.t.u-tokyo.ac.jp> > <mailto:ooura@mmm.t.u-tokyo.ac.jp <mailto:ooura@mmm.t.u-tokyo.ac.jp>>). > You may use, copy, modify this code for any purpose and > without fee. You may distribute this ORIGINAL package. > > Permission to distribute the modified gamma function code with

On 07/10/16 14:16, Jacob Vanderplas wrote: the

...

> FFTLog package has been granted (email from Takuya Ooura to

Andrew

...

> Hamilton dated 16 March 1999). > > Hence I think it is compatible with SciPy, or am I wrong? > > Alternatively it could be replaced with

scipy.special.loggamma, I

...

> think, but I do not know if it is possible to mix Fortran and

Python

...

> code with f2py. > > -- [2], [3], [4] -- > They are from the NCAR suite of FFT routines (Swarztrauber

1979).

...

> > As far as I can see they are already in SciPy, in > scipy/fftpack/src/dfftpack/ (dfftb.f, dfftf.f, and dffti.f). > > So there is definitely not a licensing problem here. > > -- [5] -- > This is the actual FFTLog routine from Hamilton. > > I asked Hamilton for permission before publishing his source

code at

...

> github.com/prisae/fftlog <http://github.com/prisae/fftlog> <http://github.com/prisae/fftlog <http://github.com/prisae/fftlog>>. > > His email response was: > > -- start email from 28/09/2016 -- > Dieter, > > You are welcome to use fftlog in any way you choose. Please

note the

...

> credits commented in the code: > > c FFTLog uses the NCAR suite of FFT routines, > c and a modified version of the complex Gamma function > c from the gamerf package at > c http://momonga.t.u-tokyo.ac.jp/~ooura/gamerf.html <http://momonga.t.u-tokyo.ac.jp/~ooura/gamerf.html> > <http://momonga.t.u-tokyo.ac.jp/~ooura/gamerf.html <http://momonga.t.u-tokyo.ac.jp/~ooura/gamerf.html>> . > c The original gamerf copyright statement states: > c Copyright(C) 1996 Takuya OOURA (email: > ooura@mmm.t.u-tokyo.ac.jp <mailto:ooura@mmm.t.u-tokyo.ac.jp> <mailto:ooura@mmm.t.u-tokyo.ac.jp <mailto:ooura@mmm.t.u-tokyo.ac.jp

...

). > c You may use, copy, modify this code for any purpose and > c without fee. You may distribute this ORIGINAL package. > c > c Permission to distribute the modified gamma function code > c with the FFTLog package has been granted > c (email from Takuya Ooura to Andrew Hamilton dated 16 March

1999).

...

> > Andrew > -- end email from 28/09/2016 -- > > _______________________________________________ > SciPy-Dev mailing list > SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org> <mailto:SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org>> > https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev> > <https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev>> > > > > > _______________________________________________ > SciPy-Dev mailing list > SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org> > https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev> > _______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org> https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev>

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org https://mail.scipy.org/mailman/listinfo/scipy-dev

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org https://mail.scipy.org/mailman/listinfo/scipy-dev

On Sat, Oct 8, 2016 at 9:11 AM, Ralf Gommers <ralf.gommers@gmail.com> wrote:

On Sat, Oct 8, 2016 at 8:59 AM, Dieter Werthmüller <dieter@werthmuller.org

...

wrote:

...

Jake,

Thanks for the clarification. I will try to get the permissions from the authors.

What is regarded as sufficient? Is an email from the author, granting me/SciPy to distribute their code with a specific, BSD-style license sufficient? Or do they necessarily have to change the websites where they host the code to include the license?

An email stating that the code can be distributed under a BSD license (or MIT or other compatible license) is enough.

Some delay on the line, missed Jake's answer. Email is enough, but a change in the repo would of course be even better. Ralf

On Sat, Oct 8, 2016 at 1:58 PM, Dieter Werthmüller <dieter@werthmuller.org> wrote:

Evening,

I wrote to Andrew Hamilton, and below is his answer. He has basically no problem with it at all, I just don't know if his response is explicit enough.

I think my email was quite extensive, and I assume that is as much as we will get from him. I also do not expect him to change his on "13 Mar 1999, 21:17" from TeX translated website...

What do you think, is this enough?

Yes, that's perfectly fine. Thanks for clarifying that Dieter. Cheers, Ralf

I also wrote to Takuya Ooura, and will let you know of his response, if I get one. However, as there are other complex logarithmic double precision gamma functions around, one already in scipy, this piece is not mission critical.

Regards, Dieter

========== START email correspondence with Andrew Hamilton ========== Subject: Re: FFTLog - license From: Andrew Hamilton <andrew.hamilton@colorado.edu> Date: 07/10/16 18:22 To: Dieter Werthmüller <dieter.werthmuller@gmx.ch> CC: Andrew.Hamilton@colorado.edu

Dieter,

I approve your adding the license language you suggest to FFTLog, and making available the resulting package for distribution.

Andrew

On 10/07/2016 03:42 PM, Dieter Werthmüller wrote:

...

Dear Andrew,

Please apologize me bothering you again.

After I published the code to wrap your FFTLog for Python I thought that it would be much better if your FFTLog would make it straight into the scientific library of Python. This would make your FFTLog available to a much wider audience.

I contacted the developers of SciPy (http://scipy.org), and they are interested in including your code. However, there is one issue: licensing. Code that is published on the web without a license file is copyrighted under law, and SciPy can for this reason not include your code into their library.

All I ask for is if you could confirm to us by email that we are allowed to distribute your FFTLog under the BSD-3-Clause license: https://opensource.org/licenses/BSD-3-Clause

The license is just a suggestion, any other BSD or MIT compatible license would be fine as well. (For the same, legal reasons we recommend to publish a license file on your website too, but that is obviously entirely up to you. It might, however, clarify things for future visitors.)

It would only affect your fftlog.f file, and the changes you made to cdgamma.f. I will write Takuya OOURA as well regarding the original cdgamma.f-file, asking him the same favour. And the three drfft*.f are already in the SciPy-library with the whole FFTPack.

If you are interested why this issue arises, Jake Vanderplas, one of the developers of SciPy, wrote an interesting article about the topic: http://www.astrobetter.com/blog/2014/03/10/the-whys-and-hows -of-licensing-scientific-code/

Thank you again for your time and for making FFTLog available! Best regards, Dieter

========== END email correspondence with Andrew Hamilton ==========

On 07/10/16 15:13, Ralf Gommers wrote:

...

On Sat, Oct 8, 2016 at 9:11 AM, Ralf Gommers <ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com>> wrote:

On Sat, Oct 8, 2016 at 8:59 AM, Dieter Werthmüller <dieter@werthmuller.org <mailto:dieter@werthmuller.org>> wrote:

Jake,

Thanks for the clarification. I will try to get the permissions from the authors.

What is regarded as sufficient? Is an email from the author, granting me/SciPy to distribute their code with a specific, BSD-style license sufficient? Or do they necessarily have to change the websites where they host the code to include the license?

An email stating that the code can be distributed under a BSD license (or MIT or other compatible license) is enough.

Some delay on the line, missed Jake's answer. Email is enough, but a change in the repo would of course be even better.

Ralf

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org https://mail.scipy.org/mailman/listinfo/scipy-dev

_______________________________________________

SciPy-Dev mailing list SciPy-Dev@scipy.org https://mail.scipy.org/mailman/listinfo/scipy-dev

Joshua Thanks for your input. How could I include the Cython loggamma function within my setup script, so that fftlog.f would pick it up? Adjusting fftlog.f to pick up dfft*.f from scipy/fftpack/src/dfftpack worked without problems. So if I could do the same for loggamma, then the only new file would be fftlog.f, for which we have the permission. Thanks, Dieter On 07/10/16 21:18, Joshua Wilson wrote:

Re log gamma: under the hood the SciPy versions are in C (real version) and Cython (complex version), so linking them up with Fortran code is definitely doable.

On Fri, Oct 7, 2016 at 8:16 PM, Ralf Gommers <ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com>> wrote:

On Sat, Oct 8, 2016 at 1:58 PM, Dieter Werthmüller <dieter@werthmuller.org <mailto:dieter@werthmuller.org>> wrote:

Evening,

I wrote to Andrew Hamilton, and below is his answer. He has basically no problem with it at all, I just don't know if his response is explicit enough.

I think my email was quite extensive, and I assume that is as much as we will get from him. I also do not expect him to change his on "13 Mar 1999, 21:17" from TeX translated website...

What do you think, is this enough?

Yes, that's perfectly fine. Thanks for clarifying that Dieter.

Cheers, Ralf

I also wrote to Takuya Ooura, and will let you know of his response, if I get one. However, as there are other complex logarithmic double precision gamma functions around, one already in scipy, this piece is not mission critical.

Regards, Dieter

========== START email correspondence with Andrew Hamilton ========== Subject: Re: FFTLog - license From: Andrew Hamilton <andrew.hamilton@colorado.edu <mailto:andrew.hamilton@colorado.edu>> Date: 07/10/16 18:22 To: Dieter Werthmüller <dieter.werthmuller@gmx.ch <mailto:dieter.werthmuller@gmx.ch>> CC: Andrew.Hamilton@colorado.edu <mailto:Andrew.Hamilton@colorado.edu>

Dieter,

I approve your adding the license language you suggest to FFTLog, and making available the resulting package for distribution.

Andrew

On 10/07/2016 03:42 PM, Dieter Werthmüller wrote: > Dear Andrew, > > Please apologize me bothering you again. > > After I published the code to wrap your FFTLog for Python I thought that > it would be much better if your FFTLog would make it straight into the > scientific library of Python. This would make your FFTLog available to a > much wider audience. > > I contacted the developers of SciPy (http://scipy.org), and they are > interested in including your code. However, there is one issue: > licensing. Code that is published on the web without a license file is > copyrighted under law, and SciPy can for this reason not include your > code into their library. > > All I ask for is if you could confirm to us by email that we are allowed > to distribute your FFTLog under the BSD-3-Clause license: > https://opensource.org/licenses/BSD-3-Clause <https://opensource.org/licenses/BSD-3-Clause> > > The license is just a suggestion, any other BSD or MIT compatible > license would be fine as well. (For the same, legal reasons we recommend > to publish a license file on your website too, but that is obviously > entirely up to you. It might, however, clarify things for future visitors.) > > It would only affect your fftlog.f file, and the changes you made to > cdgamma.f. I will write Takuya OOURA as well regarding the original > cdgamma.f-file, asking him the same favour. And the three drfft*.f are > already in the SciPy-library with the whole FFTPack. > > If you are interested why this issue arises, Jake Vanderplas, one of the > developers of SciPy, wrote an interesting article about the topic: > http://www.astrobetter.com/blog/2014/03/10/the-whys-and-hows-of-licensing-sc... <http://www.astrobetter.com/blog/2014/03/10/the-whys-and-hows-of-licensing-sc...> > > Thank you again for your time and for making FFTLog available! > Best regards, > Dieter > ========== END email correspondence with Andrew Hamilton ==========

On 07/10/16 15:13, Ralf Gommers wrote:

On Sat, Oct 8, 2016 at 9:11 AM, Ralf Gommers <ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com> <mailto:ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com>>> wrote:

On Sat, Oct 8, 2016 at 8:59 AM, Dieter Werthmüller <dieter@werthmuller.org <mailto:dieter@werthmuller.org> <mailto:dieter@werthmuller.org <mailto:dieter@werthmuller.org>>> wrote:

Jake,

Thanks for the clarification. I will try to get the permissions from the authors.

What is regarded as sufficient? Is an email from the author, granting me/SciPy to distribute their code with a specific, BSD-style license sufficient? Or do they necessarily have to change the websites where they host the code to include the license?

An email stating that the code can be distributed under a BSD license (or MIT or other compatible license) is enough.

Some delay on the line, missed Jake's answer. Email is enough, but a change in the repo would of course be even better.

Ralf

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org> https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev>

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org> https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev>

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org> https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev>

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org https://mail.scipy.org/mailman/listinfo/scipy-dev

Joshua, Thanks for the additional instructions for the fun :D Cython is at the moment unfortunately slightly outside my knowledge zone. I will therefore wait a little while to see if I hear back from Takuya Ooura regarding license. If not I might give it a try, given I find the time to dive into Cython. The actual fftlog.f-routine is not that long, and most if not all of the heavy lifting is done by rfft/irfft. Another approach might therefore be to rewrite fftlog.f into (c)python, instead of wrapping it. Just thinking. I keep you informed on the status of cdgamma.f. Dieter On 08/10/16 15:04, Joshua Wilson wrote:

Dieter,

First, a warning: this is going to involve some fun. ;-) Before going too far you should probably check that the different versions of the complex Gamma functions use the same branch cuts, etc. Note that if you want this to work in your external git repo you'll need to build against scipy master.

Steps should be roughly: -- Write a Cython script that wraps the Cython version of SciPy's loggamma. See

https://scipy.github.io/devdocs/special.cython_special.html#module-scipy.spe... <https://scipy.github.io/devdocs/special.cython_special.html#module-scipy.spe...>

for info on cimporting loggamma. -- Export the wrapper as a C function using these steps:

http://docs.cython.org/en/latest/src/userguide/external_C_code.html#using-cy... <http://docs.cython.org/en/latest/src/userguide/external_C_code.html#using-cy...>

-- Add the Cythonized C file to your setup and add an Interface block to your Fortran code letting it know how to call the C function.

On Sat, Oct 8, 2016 at 1:11 PM, Dieter Werthmüller <dieter@werthmuller.org <mailto:dieter@werthmuller.org>> wrote:

Joshua

Thanks for your input. How could I include the Cython loggamma function within my setup script, so that fftlog.f would pick it up?

Adjusting fftlog.f to pick up dfft*.f from scipy/fftpack/src/dfftpack worked without problems. So if I could do the same for loggamma, then the only new file would be fftlog.f, for which we have the permission.

Thanks, Dieter

On 07/10/16 21:18, Joshua Wilson wrote:

Re log gamma: under the hood the SciPy versions are in C (real version) and Cython (complex version), so linking them up with Fortran code is definitely doable.

On Fri, Oct 7, 2016 at 8:16 PM, Ralf Gommers <ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com> <mailto:ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com>>> wrote:

On Sat, Oct 8, 2016 at 1:58 PM, Dieter Werthmüller <dieter@werthmuller.org <mailto:dieter@werthmuller.org> <mailto:dieter@werthmuller.org <mailto:dieter@werthmuller.org>>> wrote:

Evening,

I wrote to Andrew Hamilton, and below is his answer. He has basically no problem with it at all, I just don't know if his response is explicit enough.

I think my email was quite extensive, and I assume that is as much as we will get from him. I also do not expect him to change his on "13 Mar 1999, 21:17" from TeX translated website...

What do you think, is this enough?

Yes, that's perfectly fine. Thanks for clarifying that Dieter.

Cheers, Ralf

I also wrote to Takuya Ooura, and will let you know of his response, if I get one. However, as there are other complex logarithmic double precision gamma functions around, one already in scipy, this piece is not mission critical.

Regards, Dieter

========== START email correspondence with Andrew Hamilton ========== Subject: Re: FFTLog - license From: Andrew Hamilton <andrew.hamilton@colorado.edu <mailto:andrew.hamilton@colorado.edu> <mailto:andrew.hamilton@colorado.edu <mailto:andrew.hamilton@colorado.edu>>> Date: 07/10/16 18:22 To: Dieter Werthmüller <dieter.werthmuller@gmx.ch <mailto:dieter.werthmuller@gmx.ch> <mailto:dieter.werthmuller@gmx.ch <mailto:dieter.werthmuller@gmx.ch>>> CC: Andrew.Hamilton@colorado.edu <mailto:Andrew.Hamilton@colorado.edu> <mailto:Andrew.Hamilton@colorado.edu <mailto:Andrew.Hamilton@colorado.edu>>

Dieter,

I approve your adding the license language you suggest to FFTLog, and making available the resulting package for distribution.

Andrew

On 10/07/2016 03:42 PM, Dieter Werthmüller wrote: > Dear Andrew, > > Please apologize me bothering you again. > > After I published the code to wrap your FFTLog for Python I thought that > it would be much better if your FFTLog would make it straight into the > scientific library of Python. This would make your FFTLog available to a > much wider audience. > > I contacted the developers of SciPy (http://scipy.org), and they are > interested in including your code. However, there is one issue: > licensing. Code that is published on the web without a license file is > copyrighted under law, and SciPy can for this reason not include your > code into their library. > > All I ask for is if you could confirm to us by email that we are allowed > to distribute your FFTLog under the BSD-3-Clause license: > https://opensource.org/licenses/BSD-3-Clause <https://opensource.org/licenses/BSD-3-Clause> <https://opensource.org/licenses/BSD-3-Clause <https://opensource.org/licenses/BSD-3-Clause>> > > The license is just a suggestion, any other BSD or MIT compatible > license would be fine as well. (For the same, legal reasons we recommend > to publish a license file on your website too, but that is obviously > entirely up to you. It might, however, clarify things for future visitors.) > > It would only affect your fftlog.f file, and the changes you made to > cdgamma.f. I will write Takuya OOURA as well regarding the original > cdgamma.f-file, asking him the same favour. And the three drfft*.f are > already in the SciPy-library with the whole FFTPack. > > If you are interested why this issue arises, Jake Vanderplas, one of the > developers of SciPy, wrote an interesting article about the topic: >

http://www.astrobetter.com/blog/2014/03/10/the-whys-and-hows-of-licensing-sc... <http://www.astrobetter.com/blog/2014/03/10/the-whys-and-hows-of-licensing-sc...>

<http://www.astrobetter.com/blog/2014/03/10/the-whys-and-hows-of-licensing-sc... <http://www.astrobetter.com/blog/2014/03/10/the-whys-and-hows-of-licensing-scientific-code/>> > > Thank you again for your time and for making FFTLog available! > Best regards, > Dieter > ========== END email correspondence with Andrew Hamilton ==========

On 07/10/16 15:13, Ralf Gommers wrote:

On Sat, Oct 8, 2016 at 9:11 AM, Ralf Gommers <ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com> <mailto:ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com>> <mailto:ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com> <mailto:ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com>>>> wrote:

On Sat, Oct 8, 2016 at 8:59 AM, Dieter Werthmüller <dieter@werthmuller.org <mailto:dieter@werthmuller.org> <mailto:dieter@werthmuller.org <mailto:dieter@werthmuller.org>> <mailto:dieter@werthmuller.org <mailto:dieter@werthmuller.org> <mailto:dieter@werthmuller.org <mailto:dieter@werthmuller.org>>>> wrote:

Jake,

Thanks for the clarification. I will try to get the permissions from the authors.

What is regarded as sufficient? Is an email from the author, granting me/SciPy to distribute their code with a specific, BSD-style license sufficient? Or do they necessarily have to change the websites where they host the code to include the license?

An email stating that the code can be distributed under a BSD license (or MIT or other compatible license) is enough.

Some delay on the line, missed Jake's answer. Email is enough, but a change in the repo would of course be even better.

Ralf

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org> <mailto:SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org>> https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev> <https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev>>

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org> <mailto:SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org>> https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev> <https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev>>

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org> <mailto:SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org>> https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev> <https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev>>

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org> https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev>

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org> https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev>

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org https://mail.scipy.org/mailman/listinfo/scipy-dev

On Wed, Oct 12, 2016 at 6:38 PM, Dieter Werthmüller <dieter@werthmuller.org> wrote:

Everyone,

I have a response from Takuya Ooura, appended below. He has no problem with it either, I just don't know if his response is explicit enough (again).

What do you think, is this enough?

Hmm, it's slightly ambiguous. I would tend to say no, better to get a clear statement. It sounds like he's happy, so you just need to spell out exactly what you need him to say. Or you add a BSD-3 license and "Copyright Takuya Ooura" at the top of that file in your scipy fork, point him to it, and say "can you please agree to that exact text". Cheers, Ralf

Regards, Dieter

========== START email correspondence with Takuya Ooura ========== Subject: Re: cdgamma - license From: Takuya Ooura <ooura@kurims.kyoto-u.ac.jp> Date: Wed, 12 Oct 2016 21:31:01 +0900 (JST) To: dieter@werthmuller.org

Dear Dieter Werthm üller,

Please use the modified version of cdgamma.f. The license of my code is similar to the BSD-3-Clause license at present.

-- Takuya Ooura Email : ooura@kurims.kyoto-u.ac.jp

...

Dear Takuya Ooura,

Firstly thank you very much for making your code available on your website.

I am writing to you regarding your cdgamma.f function. Andrew Hamilton used a modified version of it for his FFTLog, and got your written permission to distribute the modified version in his code.

We would like to include FFTLog in the scientific library of the python programming language, SciPy, and with it the modified version of cdgamma.f.

The copyright statement distributed with your code states

Copyright(C) 1996 Takuya OOURA (email: ooura@mmm.t.u-tokyo.ac.jp). You may use, copy, modify this code for any purpose and without fee. You may distribute this ORIGINAL package.

which makes it impossible for SciPy to include the modified version of cdgamma.f in FFTLog into their library, as the version in FFTLog is a MODIFIED version, and not the original version.

Would it be possible that you could give SciPy the permission to distribute cdgamma.f under a permissive license? My suggestion would be the BSD-3-Clause license, https://opensource.org/licenses/BSD-3-Clause But any other BSD or MIT compatible license would be fine as well.

Thank you for your time and for making cdgamma.f available on your website.

Best regards, Dieter Werthm üller

========== END email correspondence with Takuya Ooura ==========

On 08/10/16 15:04, Joshua Wilson wrote:

...

Dieter,

First, a warning: this is going to involve some fun. ;-) Before going too far you should probably check that the different versions of the complex Gamma functions use the same branch cuts, etc. Note that if you want this to work in your external git repo you'll need to build against scipy master.

Steps should be roughly: -- Write a Cython script that wraps the Cython version of SciPy's loggamma. See

https://scipy.github.io/devdocs/special.cython_special.html# module-scipy.special.cython_special <https://scipy.github.io/devdocs/special.cython_special. html#module-scipy.special.cython_special>

for info on cimporting loggamma. -- Export the wrapper as a C function using these steps:

http://docs.cython.org/en/latest/src/userguide/external_C_ code.html#using-cython-declarations-from-c <http://docs.cython.org/en/latest/src/userguide/external_C_ code.html#using-cython-declarations-from-c>

-- Add the Cythonized C file to your setup and add an Interface block to your Fortran code letting it know how to call the C function.

On Sat, Oct 8, 2016 at 1:11 PM, Dieter Werthmüller <dieter@werthmuller.org <mailto:dieter@werthmuller.org>> wrote:

Joshua

Thanks for your input. How could I include the Cython loggamma function within my setup script, so that fftlog.f would pick it up?

Adjusting fftlog.f to pick up dfft*.f from scipy/fftpack/src/dfftpack worked without problems. So if I could do the same for loggamma, then the only new file would be fftlog.f, for which we have the permission.

Thanks, Dieter

On 07/10/16 21:18, Joshua Wilson wrote:

Re log gamma: under the hood the SciPy versions are in C (real version) and Cython (complex version), so linking them up with Fortran code is definitely doable.

On Fri, Oct 7, 2016 at 8:16 PM, Ralf Gommers <ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com> <mailto:ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com>>> wrote:

On Sat, Oct 8, 2016 at 1:58 PM, Dieter Werthmüller <dieter@werthmuller.org <mailto:dieter@werthmuller.org> <mailto:dieter@werthmuller.org <mailto:dieter@werthmuller.org>>> wrote:

Evening,

I wrote to Andrew Hamilton, and below is his answer. He has basically no problem with it at all, I just don't know if his response is explicit enough.

I think my email was quite extensive, and I assume that is as much as we will get from him. I also do not expect him to change his on "13 Mar 1999, 21:17" from TeX translated website...

What do you think, is this enough?

Yes, that's perfectly fine. Thanks for clarifying that Dieter.

Cheers, Ralf

I also wrote to Takuya Ooura, and will let you know of his response, if I get one. However, as there are other complex logarithmic double precision gamma functions around, one already in scipy, this piece is not mission critical.

Regards, Dieter

========== START email correspondence with Andrew Hamilton ========== Subject: Re: FFTLog - license From: Andrew Hamilton <andrew.hamilton@colorado.edu <mailto:andrew.hamilton@colorado.edu> <mailto:andrew.hamilton@colorado.edu <mailto:andrew.hamilton@colorado.edu>>> Date: 07/10/16 18:22 To: Dieter Werthmüller <dieter.werthmuller@gmx.ch <mailto:dieter.werthmuller@gmx.ch> <mailto:dieter.werthmuller@gmx.ch <mailto:dieter.werthmuller@gmx.ch>>> CC: Andrew.Hamilton@colorado.edu <mailto:Andrew.Hamilton@colorado.edu> <mailto:Andrew.Hamilton@colorado.edu

<mailto:Andrew.Hamilton@colorado.edu>>

Dieter,

I approve your adding the license language you suggest to FFTLog, and making available the resulting package for distribution.

Andrew

On 10/07/2016 03:42 PM, Dieter Werthmüller wrote: > Dear Andrew, > > Please apologize me bothering you again. > > After I published the code to wrap your FFTLog for Python I thought that > it would be much better if your FFTLog would make it straight into the > scientific library of Python. This would make your FFTLog available to a > much wider audience. > > I contacted the developers of SciPy (http://scipy.org), and they are > interested in including your code. However, there is one issue: > licensing. Code that is published on the web without a license file is > copyrighted under law, and SciPy can for this reason not include your > code into their library. > > All I ask for is if you could confirm to us by email that we are allowed > to distribute your FFTLog under the BSD-3-Clause license: > https://opensource.org/licenses/BSD-3-Clause <https://opensource.org/licenses/BSD-3-Clause> <https://opensource.org/licenses/BSD-3-Clause <https://opensource.org/licenses/BSD-3-Clause>> > > The license is just a suggestion, any other BSD or MIT compatible > license would be fine as well. (For the same, legal reasons we recommend > to publish a license file on your website too, but that is obviously > entirely up to you. It might, however, clarify things for future visitors.) > > It would only affect your fftlog.f file, and the changes you made to > cdgamma.f. I will write Takuya OOURA as well regarding the original > cdgamma.f-file, asking him the same favour. And the three drfft*.f are > already in the SciPy-library with the whole FFTPack. > > If you are interested why this issue arises, Jake Vanderplas, one of the > developers of SciPy, wrote an interesting article about the topic: >

http://www.astrobetter.com/blog/2014/03/10/the-whys-and-hows -of-licensing-scientific-code/ <http://www.astrobetter.com/blog/2014/03/10/the-whys-and-how s-of-licensing-scientific-code/>

<http://www.astrobetter.com/blog/2014/03/10/the-whys-and-how s-of-licensing-scientific-code/ <http://www.astrobetter.com/blog/2014/03/10/the-whys-and-how s-of-licensing-scientific-code/>> > > Thank you again for your time and for making FFTLog available! > Best regards, > Dieter > ========== END email correspondence with Andrew Hamilton ==========

On 07/10/16 15:13, Ralf Gommers wrote:

On Sat, Oct 8, 2016 at 9:11 AM, Ralf Gommers <ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com> <mailto:ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com>> <mailto:ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com> <mailto:ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com>>>> wrote:

On Sat, Oct 8, 2016 at 8:59 AM, Dieter Werthmüller <dieter@werthmuller.org <mailto:dieter@werthmuller.org> <mailto:dieter@werthmuller.org <mailto:dieter@werthmuller.org>> <mailto:dieter@werthmuller.org <mailto:dieter@werthmuller.org> <mailto:dieter@werthmuller.org <mailto:dieter@werthmuller.org>>>> wrote:

Jake,

Thanks for the clarification. I will try to get the permissions from the authors.

What is regarded as sufficient? Is an email from the author, granting me/SciPy to distribute their code with a specific, BSD-style license sufficient? Or do they necessarily have to change the websites where they host the code to include the license?

An email stating that the code can be distributed under a BSD license (or MIT or other compatible license) is enough.

Some delay on the line, missed Jake's answer. Email is enough, but a change in the repo would of course be even better.

Ralf

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org> <mailto:SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org>> https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev> <https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev>>

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org> <mailto:SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org>> https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev> <https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev>>

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org> <mailto:SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org>> https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev> <https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev>>

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org> https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev>

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org> https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev>

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org https://mail.scipy.org/mailman/listinfo/scipy-dev

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org https://mail.scipy.org/mailman/listinfo/scipy-dev

I agree, probably the best would be to use the already existing FFTPACK and the already existing complex gamma function, which would mean only one file, fftlog.f, has to be added. In this regard, it might be best to first ask if there is sufficient interest for a logarithmic FFT, as you point out Joshua, before investing more time. Any opinions? For anyone interested I list the website of the original FFTLog again: http://casa.colorado.edu/~ajsh/FFTLog Ralf, one question out of interest and to learn something about licenses: Takuya's license has 3 sentences: 1. Copyright(C) 1996 Takuya OOURA (email: ooura@mmm.t.u-tokyo.ac.jp). 2. You may use, copy, modify this code for any purpose and without fee. 3. You may distribute this ORIGINAL package. In the second sentence, it basically states that you can do whatever you want with it. Then as far as I understand the third sentence, it just gives additionally and specifically the permission to even distribute the unchanged package. Or does the absence of the word 'distribute' in the second sentence cause a problem? Thank you both for your continued inputs, Dieter On 23/10/16 22:32, Joshua Wilson wrote:

TBH I'd rather we try to use SciPy's version anyway rather than have an implementation of unknown quality sitting around in the wrong module. It people want this feature (I don't know enough about it to have an opinion), then I can send a PR to your fork with necessary changes to special. I *think* this can be done in a portable way even though we can't use ISO_C_BINDING. (Or can we? It's from the 2003 standard but gfortran and ifort support it.)

On Sun, Oct 23, 2016 at 3:00 AM, Ralf Gommers <ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com>> wrote:

On Wed, Oct 12, 2016 at 6:38 PM, Dieter Werthmüller <dieter@werthmuller.org <mailto:dieter@werthmuller.org>> wrote:

Everyone,

I have a response from Takuya Ooura, appended below. He has no problem with it either, I just don't know if his response is explicit enough (again).

What do you think, is this enough?

Hmm, it's slightly ambiguous. I would tend to say no, better to get a clear statement. It sounds like he's happy, so you just need to spell out exactly what you need him to say. Or you add a BSD-3 license and "Copyright Takuya Ooura" at the top of that file in your scipy fork, point him to it, and say "can you please agree to that exact text".

Cheers, Ralf

Regards, Dieter

========== START email correspondence with Takuya Ooura ========== Subject: Re: cdgamma - license From: Takuya Ooura <ooura@kurims.kyoto-u.ac.jp <mailto:ooura@kurims.kyoto-u.ac.jp>> Date: Wed, 12 Oct 2016 21:31:01 +0900 (JST) To: dieter@werthmuller.org <mailto:dieter@werthmuller.org>

Dear Dieter Werthm üller,

Please use the modified version of cdgamma.f. The license of my code is similar to the BSD-3-Clause license at present.

-- Takuya Ooura Email : ooura@kurims.kyoto-u.ac.jp <mailto:ooura@kurims.kyoto-u.ac.jp>

>Dear Takuya Ooura, > >Firstly thank you very much for making your code available on your website. > >I am writing to you regarding your cdgamma.f function. Andrew Hamilton >used a modified version of it for his FFTLog, and got your written >permission to distribute the modified version in his code. > >We would like to include FFTLog in the scientific library of the python >programming language, SciPy, and with it the modified version of cdgamma.f. > >The copyright statement distributed with your code states > > Copyright(C) 1996 Takuya OOURA (email: ooura@mmm.t.u-tokyo.ac.jp <mailto:ooura@mmm.t.u-tokyo.ac.jp>). > You may use, copy, modify this code for any purpose and > without fee. You may distribute this ORIGINAL package. > >which makes it impossible for SciPy to include the modified version of >cdgamma.f in FFTLog into their library, as the version in FFTLog is a >MODIFIED version, and not the original version. > >Would it be possible that you could give SciPy the permission to >distribute cdgamma.f under a permissive license? My suggestion would be >the BSD-3-Clause license, >https://opensource.org/licenses/BSD-3-Clause <https://opensource.org/licenses/BSD-3-Clause> >But any other BSD or MIT compatible license would be fine as well. > >Thank you for your time and for making cdgamma.f available on your website. > >Best regards, >Dieter Werthm üller

========== END email correspondence with Takuya Ooura ==========

On 08/10/16 15:04, Joshua Wilson wrote:

Dieter,

First, a warning: this is going to involve some fun. ;-) Before going too far you should probably check that the different versions of the complex Gamma functions use the same branch cuts, etc. Note that if you want this to work in your external git repo you'll need to build against scipy master.

Steps should be roughly: -- Write a Cython script that wraps the Cython version of SciPy's loggamma. See

https://scipy.github.io/devdocs/special.cython_special.html#module-scipy.spe... <https://scipy.github.io/devdocs/special.cython_special.html#module-scipy.spe...> <https://scipy.github.io/devdocs/special.cython_special.html#module-scipy.spe... <https://scipy.github.io/devdocs/special.cython_special.html#module-scipy.special.cython_special>>

for info on cimporting loggamma. -- Export the wrapper as a C function using these steps:

http://docs.cython.org/en/latest/src/userguide/external_C_code.html#using-cy... <http://docs.cython.org/en/latest/src/userguide/external_C_code.html#using-cy...> <http://docs.cython.org/en/latest/src/userguide/external_C_code.html#using-cy... <http://docs.cython.org/en/latest/src/userguide/external_C_code.html#using-cython-declarations-from-c>>

-- Add the Cythonized C file to your setup and add an Interface block to your Fortran code letting it know how to call the C function.

On Sat, Oct 8, 2016 at 1:11 PM, Dieter Werthmüller <dieter@werthmuller.org <mailto:dieter@werthmuller.org> <mailto:dieter@werthmuller.org <mailto:dieter@werthmuller.org>>> wrote:

Joshua

Thanks for your input. How could I include the Cython loggamma function within my setup script, so that fftlog.f would pick it up?

Adjusting fftlog.f to pick up dfft*.f from scipy/fftpack/src/dfftpack worked without problems. So if I could do the same for loggamma, then the only new file would be fftlog.f, for which we have the permission.

Thanks, Dieter

On 07/10/16 21:18, Joshua Wilson wrote:

Re log gamma: under the hood the SciPy versions are in C (real version) and Cython (complex version), so linking them up with Fortran code is definitely doable.

On Fri, Oct 7, 2016 at 8:16 PM, Ralf Gommers <ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com> <mailto:ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com>> <mailto:ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com> <mailto:ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com>>>> wrote:

On Sat, Oct 8, 2016 at 1:58 PM, Dieter Werthmüller <dieter@werthmuller.org <mailto:dieter@werthmuller.org> <mailto:dieter@werthmuller.org <mailto:dieter@werthmuller.org>> <mailto:dieter@werthmuller.org <mailto:dieter@werthmuller.org> <mailto:dieter@werthmuller.org <mailto:dieter@werthmuller.org>>>> wrote:

Evening,

I wrote to Andrew Hamilton, and below is his answer. He has basically no problem with it at all, I just don't know if his response is explicit enough.

I think my email was quite extensive, and I assume that is as much as we will get from him. I also do not expect him to change his on "13 Mar 1999, 21:17" from TeX translated website...

What do you think, is this enough?

Yes, that's perfectly fine. Thanks for clarifying that Dieter.

Cheers, Ralf

I also wrote to Takuya Ooura, and will let you know of his response, if I get one. However, as there are other complex logarithmic double precision gamma functions around, one already in scipy, this piece is not mission critical.

Regards, Dieter

========== START email correspondence with Andrew Hamilton ========== Subject: Re: FFTLog - license From: Andrew Hamilton <andrew.hamilton@colorado.edu <mailto:andrew.hamilton@colorado.edu> <mailto:andrew.hamilton@colorado.edu <mailto:andrew.hamilton@colorado.edu>> <mailto:andrew.hamilton@colorado.edu <mailto:andrew.hamilton@colorado.edu> <mailto:andrew.hamilton@colorado.edu <mailto:andrew.hamilton@colorado.edu>>>> Date: 07/10/16 18:22 To: Dieter Werthmüller <dieter.werthmuller@gmx.ch <mailto:dieter.werthmuller@gmx.ch> <mailto:dieter.werthmuller@gmx.ch <mailto:dieter.werthmuller@gmx.ch>> <mailto:dieter.werthmuller@gmx.ch <mailto:dieter.werthmuller@gmx.ch> <mailto:dieter.werthmuller@gmx.ch <mailto:dieter.werthmuller@gmx.ch>>>> CC: Andrew.Hamilton@colorado.edu <mailto:Andrew.Hamilton@colorado.edu> <mailto:Andrew.Hamilton@colorado.edu <mailto:Andrew.Hamilton@colorado.edu>> <mailto:Andrew.Hamilton@colorado.edu <mailto:Andrew.Hamilton@colorado.edu>

<mailto:Andrew.Hamilton@colorado.edu <mailto:Andrew.Hamilton@colorado.edu>>>

Dieter,

I approve your adding the license language you suggest to FFTLog, and making available the resulting package for distribution.

Andrew

On 10/07/2016 03:42 PM, Dieter Werthmüller wrote: > Dear Andrew, > > Please apologize me bothering you again. > > After I published the code to wrap your FFTLog for Python I thought that > it would be much better if your FFTLog would make it straight into the > scientific library of Python. This would make your FFTLog available to a > much wider audience. > > I contacted the developers of SciPy (http://scipy.org), and they are > interested in including your code. However, there is one issue: > licensing. Code that is published on the web without a license file is > copyrighted under law, and SciPy can for this reason not include your > code into their library. > > All I ask for is if you could confirm to us by email that we are allowed > to distribute your FFTLog under the BSD-3-Clause license: > https://opensource.org/licenses/BSD-3-Clause <https://opensource.org/licenses/BSD-3-Clause> <https://opensource.org/licenses/BSD-3-Clause <https://opensource.org/licenses/BSD-3-Clause>>

<https://opensource.org/licenses/BSD-3-Clause <https://opensource.org/licenses/BSD-3-Clause> <https://opensource.org/licenses/BSD-3-Clause <https://opensource.org/licenses/BSD-3-Clause>>> > > The license is just a suggestion, any other BSD or MIT compatible > license would be fine as well. (For the same, legal reasons we recommend > to publish a license file on your website too, but that is obviously > entirely up to you. It might, however, clarify things for future visitors.) > > It would only affect your fftlog.f file, and the changes you made to > cdgamma.f. I will write Takuya OOURA as well regarding the original > cdgamma.f-file, asking him the same favour. And the three drfft*.f are > already in the SciPy-library with the whole FFTPack. > > If you are interested why this issue arises, Jake Vanderplas, one of the > developers of SciPy, wrote an interesting article about the topic: >

http://www.astrobetter.com/blog/2014/03/10/the-whys-and-hows-of-licensing-sc... <http://www.astrobetter.com/blog/2014/03/10/the-whys-and-hows-of-licensing-sc...>

<http://www.astrobetter.com/blog/2014/03/10/the-whys-and-hows-of-licensing-sc... <http://www.astrobetter.com/blog/2014/03/10/the-whys-and-hows-of-licensing-scientific-code/>>

<http://www.astrobetter.com/blog/2014/03/10/the-whys-and-hows-of-licensing-sc... <http://www.astrobetter.com/blog/2014/03/10/the-whys-and-hows-of-licensing-sc...>

<http://www.astrobetter.com/blog/2014/03/10/the-whys-and-hows-of-licensing-sc... <http://www.astrobetter.com/blog/2014/03/10/the-whys-and-hows-of-licensing-scientific-code/>>> > > Thank you again for your time and for making FFTLog available! > Best regards, > Dieter > ========== END email correspondence with Andrew Hamilton ==========

On 07/10/16 15:13, Ralf Gommers wrote:

On Sat, Oct 8, 2016 at 9:11 AM, Ralf Gommers <ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com> <mailto:ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com>> <mailto:ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com> <mailto:ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com>>> <mailto:ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com> <mailto:ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com>> <mailto:ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com> <mailto:ralf.gommers@gmail.com <mailto:ralf.gommers@gmail.com>>>>> wrote:

On Sat, Oct 8, 2016 at 8:59 AM, Dieter Werthmüller <dieter@werthmuller.org <mailto:dieter@werthmuller.org> <mailto:dieter@werthmuller.org <mailto:dieter@werthmuller.org>> <mailto:dieter@werthmuller.org <mailto:dieter@werthmuller.org> <mailto:dieter@werthmuller.org <mailto:dieter@werthmuller.org>>> <mailto:dieter@werthmuller.org <mailto:dieter@werthmuller.org> <mailto:dieter@werthmuller.org <mailto:dieter@werthmuller.org>> <mailto:dieter@werthmuller.org <mailto:dieter@werthmuller.org> <mailto:dieter@werthmuller.org <mailto:dieter@werthmuller.org>>>>> wrote:

Jake,

Thanks for the clarification. I will try to get the permissions from the authors.

What is regarded as sufficient? Is an email from the author, granting me/SciPy to distribute their code with a specific, BSD-style license sufficient? Or do they necessarily have to change the websites where they host the code to include the license?

An email stating that the code can be distributed under a BSD license (or MIT or other compatible license) is enough.

Some delay on the line, missed Jake's answer. Email is enough, but a change in the repo would of course be even better.

Ralf

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org> <mailto:SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org>> <mailto:SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org> <mailto:SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org>>>

https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev> <https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev>>

<https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev> <https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev>>>

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org> <mailto:SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org>> <mailto:SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org> <mailto:SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org>>>

https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev> <https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev>>

<https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev> <https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev>>>

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org> <mailto:SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org>> <mailto:SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org> <mailto:SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org>>>

https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev> <https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev>>

<https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev> <https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev>>>

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org> <mailto:SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org>> https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev> <https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev>>

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org> <mailto:SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org>> https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev> <https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev>>

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org> https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev>

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org> https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev>

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org <mailto:SciPy-Dev@scipy.org> https://mail.scipy.org/mailman/listinfo/scipy-dev <https://mail.scipy.org/mailman/listinfo/scipy-dev>

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org https://mail.scipy.org/mailman/listinfo/scipy-dev

On Tue, Oct 25, 2016 at 3:57 AM, Daπid <davidmenhur@gmail.com> wrote:

On 24 October 2016 at 16:25, Dieter Werthmüller <dieter@werthmuller.org> wrote:

...

1. Copyright(C) 1996 Takuya OOURA (email: ooura@mmm.t.u-tokyo.ac.jp). 2. You may use, copy, modify this code for any purpose and without fee. 3. You may distribute this ORIGINAL package. In the second sentence, it basically states that you can do whatever you want with it. Then as far as I understand the third sentence, it just gives additionally and specifically the permission to even distribute the unchanged package. Or does the absence of the word 'distribute' in the second sentence cause a problem?

With my scientist hat on, I think that he intended to let you do whatever you want with it, but with my not-really-a-lawyer hat, it can be interpreted both ways, so an actual lawyer may get squeamish without a more explicit agreement.

Agreed. With common sense it's clearly fine to reuse his code, but legal stuff isn't common sense - it could also be read as giving permission to redistribute only the original code and not a modified version. Ralf

Great. I am out of the office most of November. I will pick it up latest in December again and see how far I get with my knowledge, and come back for help if I am lost. Dieter On 31/10/16 03:59, Ralf Gommers wrote:

On Tue, Oct 25, 2016 at 3:25 AM, Dieter Werthmüller <dieter@werthmuller.org <mailto:dieter@werthmuller.org>> wrote:

I agree, probably the best would be to use the already existing FFTPACK and the already existing complex gamma function, which would mean only one file, fftlog.f, has to be added.

In this regard, it might be best to first ask if there is sufficient interest for a logarithmic FFT, as you point out Joshua, before investing more time. Any opinions? For anyone interested I list the website of the original FFTLog again: http://casa.colorado.edu/~ajsh/FFTLog <http://casa.colorado.edu/%7Eajsh/FFTLog>

I haven't come across a use case for this myself, but the original paper [1] is highly cited and the citation come from a range of fields. The one fftlog.f file doesn't look worse than the average Fortran file, so this will likely be okay to maintain. So I'm fine with adding this feature.

Ralf

[1] http://www.sciencedirect.com/science/article/pii/0021999178901079

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@scipy.org https://mail.scipy.org/mailman/listinfo/scipy-dev