[Scikit-learn-general] scikit-learn in Julia

[Cedric St-Jean]

Hi,

I am working on a library to bring the scikit-learn interface to Julia (
https://github.com/cstjean/ScikitLearn.jl). It's mostly a wrapper around
scikit-learn, but I had to translate some code to make it possible to
define new estimators in Julia. Here's a translated example:
https://github.com/cstjean/ScikitLearn.jl/blob/master/examples/Pipeline_PCA_Logistic.ipynb

I would like to call it ScikitLearn.jl. Can I?

Also, can I assume that the scikit-learn documentation is covered by the
BSD license? The blurb at the bottom of each scikit-learn.org page suggests
that it is, but it's not entirely clear.

Best,

Cédric

[Andreas Mueller]

Hi Cedric.
I'm not sure about the naming of the package. As long as there is no
active involvement of the scikit-learn developers in this
package, I'd rather not have "our" name on it.
Gael?

The docs should be bsd-licensed (though I'm not sure that it the best
license for docs).
The inline docs certainly are. The user guide is a bit more ambiguous, I
agree (maybe we should cc license it?).
I think it's fair to assume that it's BSD licensed but IANAL.

I assume you follow this thread:
https://github.com/JuliaStats/Roadmap.jl/issues/11

There was a relatively complete wrapper here:
https://github.com/Rory-Finnegan/Learn.jl
I don't know where it went.

Best,
Andy

Post by Cedric St-Jean
Hi,
I am working on a library to bring the scikit-learn interface to Julia
(https://github.com/cstjean/ScikitLearn.jl). It's mostly a wrapper
around scikit-learn, but I had to translate some code to make it
possible to define new estimators in Julia. Here's a translated
https://github.com/cstjean/ScikitLearn.jl/blob/master/examples/Pipeline_PCA_Logistic.ipynb
I would like to call it ScikitLearn.jl. Can I?
Also, can I assume that the scikit-learn documentation is covered by
the BSD license? The blurb at the bottom of each scikit-learn.org
<http://scikit-learn.org> page suggests that it is, but it's not
entirely clear.
Best,
Cédric
------------------------------------------------------------------------------
_______________________________________________
Scikit-learn-general mailing list
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general

[Gael Varoquaux]

I'm not sure about the naming of the package. As long as there is no active
involvement of the scikit-learn developers in this
package, I'd rather not have "our" name on it.
Gael?

I don't know about this. I see the point of naming it scikit-learn, but I
worry like Andy of the overhead in support. How about naming it something
like Juliasklearn, it would be quite explicit that it's different, yet
related.

The docs should be bsd-licensed (though I'm not sure that it the best license
for docs).
The inline docs certainly are. The user guide is a bit more ambiguous, I agree
(maybe we should cc license it?).
I think it's fair to assume that it's BSD licensed but IANAL.

I would say that it is BSD licensed, as it is commited to the same repo.

G

[Kyle Kastner]

Is julia-learn a thing already? Juliasklearn seems a bit overloaded to me,
but naming things is hard.

On Mon, Mar 7, 2016 at 11:02 AM, Gael Varoquaux <

Post by Andreas Mueller

Post by Andreas Mueller
I'm not sure about the naming of the package. As long as there is no

active

Post by Andreas Mueller
involvement of the scikit-learn developers in this
package, I'd rather not have "our" name on it.
Gael?

I don't know about this. I see the point of naming it scikit-learn, but I
worry like Andy of the overhead in support. How about naming it something
like Juliasklearn, it would be quite explicit that it's different, yet
related.

Post by Andreas Mueller
The docs should be bsd-licensed (though I'm not sure that it the best

license

Post by Andreas Mueller
for docs).
The inline docs certainly are. The user guide is a bit more ambiguous, I

agree

Post by Andreas Mueller
(maybe we should cc license it?).
I think it's fair to assume that it's BSD licensed but IANAL.

I would say that it is BSD licensed, as it is commited to the same repo.
G
------------------------------------------------------------------------------
Transform Data into Opportunity.
Accelerate data analysis in your applications with
Intel Data Analytics Acceleration Library.
Click to learn more.
http://makebettercode.com/inteldaal-eval
_______________________________________________
Scikit-learn-general mailing list
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general

[Cedric St-Jean]

Gael, Andreas,

Thank you for your comments.

Post by Andreas Mueller
I'm not sure about the naming of the package. As long as there is no
active involvement of the scikit-learn developers in this
package, I'd rather not have "our" name on it.

I understand the concern.

There's a few machine learning packages that have attempted to define an
interface, but none of them have gathered any steam so far, and navigating
the ecosystem is painful. It's as much of a social problem as a technical
problem, and by broadcasting "This is the scikit-learn interface, I changed
it as little as possible", it has better chances of succeeding.

Post by Andreas Mueller
Juliasklearn

I'm personally fine with that name, but the official guidelines (
http://docs.julialang.org/en/release-0.4/manual/packages/#guidelines-for-naming-a-package)
for package naming say:

"Avoid using Julia in your package name. It is usually clear from context
and to your users that the package is a Julia package."

"Packages that wrap external libraries or programs should be named after
those libraries or programs."

There is already Pandas.jl, Stan.jl, MATLAB.jl and Bokeh.jl following that
trend.

Maybe... NotScikitLearn.jl? ScikitForget.jl? :) It's tongue-in-cheek, but
IMO it communicates the right idea (like IceWeasel did)

Post by Andreas Mueller
I worry like Andy of the overhead in support

What are you referring to? Do you mean that people might mistakenly file
issues with scikit-learn when they have problems in ScikitLearn.jl?

Best,

Cedric

[Andreas Mueller]

Post by Cedric St-Jean
There is already Pandas.jl, Stan.jl, MATLAB.jl and Bokeh.jl following
that trend.

That is interesting. Were they done by people associated with the
original projects?
MATLAB.jl ? And mathworks was fine with that?

Post by Cedric St-Jean
Maybe... NotScikitLearn.jl? ScikitForget.jl? :) It's tongue-in-cheek,
but IMO it communicates the right idea (like IceWeasel did)

Post by Andreas Mueller
I worry like Andy of the overhead in support

What are you referring to? Do you mean that people might mistakenly
file issues with scikit-learn when they have problems in ScikitLearn.jl?

I believe that is what he means.

[Cedric St-Jean]

Post by Andreas Mueller

Post by Cedric St-Jean
There is already Pandas.jl, Stan.jl, MATLAB.jl and Bokeh.jl following
that trend.

That is interesting. Were they done by people associated with the
original projects?

As far as I can tell, no, they weren't. Stan.jl and Bokeh.jl are now both
recognized (but not explicitly supported) by their parent projects.

Post by Andreas Mueller
MATLAB.jl ? And mathworks was fine with that?

I don't know. MATLAB.jl's README has some very clear language that it's not
an official MathWorks product. I have some too, but I can step it up.

[Andreas Mueller]

Post by Cedric St-Jean

Post by Andreas Mueller

Post by Cedric St-Jean
There is already Pandas.jl, Stan.jl, MATLAB.jl and Bokeh.jl following
that trend.

That is interesting. Were they done by people associated with the
original projects?

As far as I can tell, no, they weren't. Stan.jl and Bokeh.jl are now
both recognized (but not explicitly supported) by their parent projects.

Post by Andreas Mueller
MATLAB.jl ? And mathworks was fine with that?

I don't know. MATLAB.jl's README has some very clear language that
it's not an official MathWorks product. I have some too, but I can
step it up.

Pretty sure mathworks just hasn't seen it yet / thinks it's to
insignificant to sue. IANAL but that seems like a pretty clear trademark
violation.

[Cedric St-Jean]

Here is a list of Julia packages that reference other work and start with A:

AWS.jl, AWSEC2.jl, Accumulo.jl, AmplNLWriter.jl, AppleAccelerate.jl,
ArcadeLearningEnvironment.jl, Arduino.jl, Atom.jl, Augur.jl

Full listing here: http://pkg.julialang.org/

I'm sure that some of these will get C&D letters eventually, but I'm
pointing out that it's a very entrenched convention that Julia users know
about. Julia is built around github, and people should know to post their
issues on ScikitLearn.jl, because that's how everybody does it.

Post by Cedric St-Jean

Post by Andreas Mueller

Post by Cedric St-Jean
There is already Pandas.jl, Stan.jl, MATLAB.jl and Bokeh.jl following
that trend.

That is interesting. Were they done by people associated with the
original projects?

As far as I can tell, no, they weren't. Stan.jl and Bokeh.jl are now both
recognized (but not explicitly supported) by their parent projects.

Post by Andreas Mueller
MATLAB.jl ? And mathworks was fine with that?

I don't know. MATLAB.jl's README has some very clear language that it's
not an official MathWorks product. I have some too, but I can step it up.