Ok, now I'm confused.
For future reference, you call a 2D function a function that is defined
> over the 2 dimensional space e.g. has 2 variables,
>
Correct, and that would be f = f(x,y), exactly what I'm looking for (a 2D
regression).
> hence the example in the page is a 2D example.
>
No, the
On Wed, Jan 23, 2013 at 01:57:14AM -0500, Ronnie Ghose wrote:
> haha should i try making one? ;)
If you feel up to the challenge! Remember: heavily tested is important.
Also, it should be fast, as joblib is about performance, but shouldn't
have compiled code, as joblib does not include compiled co
haha should i try making one? ;)
On Wed, Jan 23, 2013 at 1:44 AM, Gael Varoquaux <
gael.varoqu...@normalesup.org> wrote:
> On Wed, Jan 23, 2013 at 01:39:28AM -0500, Ronnie Ghose wrote:
> > why? if it's binary we can just add error correcting codes no?
>
> Good answer :). Everything is always eas
I personally feel that the term `affiliated project` is a bit too
strong. It implicitly means that we endorse the project, some kind of
quality guarantee. Do we really want to take that responsibility? This
may end up being a burden for us. Therefore, I think I prefer the term
`related project`.
B
On Wed, Jan 23, 2013 at 01:39:28AM -0500, Ronnie Ghose wrote:
> why? if it's binary we can just add error correcting codes no?
Good answer :). Everything is always easy in theory, but implementation
can often be more tricky. Do we have a robust and fast implementation of
error correcting codes in
why? if it's binary we can just add error correcting codes no?
On Wed, Jan 23, 2013 at 1:34 AM, Gael Varoquaux <
gael.varoqu...@normalesup.org> wrote:
> On Tue, Jan 22, 2013 at 06:29:26PM -0500, Ronnie Ghose wrote:
> > Any point in adding data redundancy to joblib dumped objects?
>
> This is not
On Wed, Jan 23, 2013 at 12:16:32AM +, Afik Cohen wrote:
Hi, I'm working with Ark on this project. Yes, that's what it looks like
> - some investigation into this appears to show that either this is a bug
> in zlib (the length returned is incorrect) or this is a bug in joblib.dump
> (when cal
On Tue, Jan 22, 2013 at 06:29:26PM -0500, Ronnie Ghose wrote:
> Any point in adding data redundancy to joblib dumped objects?
This is not easy at all.
Gaël
--
Master Visual Studio, SharePoint, SQL, ASP.NET, C# 2012, HTML
Ahh ok,
For future reference, you call a 2D function a function that is defined
over the 2 dimensional space e.g. has 2 variables, hence the example in the
page is a 2D example. What you need would be a 3D function. In Machine
Learning, we usually call these independent dimensions features. So you
If I'm not wrong, this is fitting a model of the form f(x) = y (univariate
regression, 1d). I want to fit a model of the form f(x,y) = z (bivariate
regression, 2d).
Cheers,
-fernando
Sent from my iPhone
On Jan 22, 2013, at 9:31 PM, Leon Palafox wrote:
This example is a 2D regression isn't it?
This example is a 2D regression isn't it?
http://scikit-learn.org/dev/modules/gaussian_process.html#an-introductory-regression-example
On Wed, Jan 23, 2013 at 2:25 PM, Fernando Paolo wrote:
> Hello,
>
> I'm new to scikit-learn, and wonder if it is possible to perform 2D
> Gaussian Process Re
Hello,
I'm new to scikit-learn, and wonder if it is possible to perform 2D
Gaussian Process Regression? In other words, I would like to perform a 2d
kriging interpolation of spatial data (e.g., some geophysical field), where
each data point is represented as (x,y,z) = (lon,lat,var). I can't figure
+1 for affiliated package, which I have wished for a long time :)
If I understand it right, it is like a contrib repo for R? I think I will
be a great place for some research codes and state-of-the-art but not time
tested algorithms. And move to upstream will be another great idea if we
have clean
Gael Varoquaux writes:
>
> On Tue, Jan 22, 2013 at 10:30:01PM +, Ark wrote:
> > /home/n7/newenv/lib/python2.6/site-
> > packages/sklearn/externals/joblib/numpy_pickle.pyc in
read_zfile(file_handle)
> > 69 assert len(data) == length, (
> > 70 "Incorrect data length whil
Ronnie Ghose writes:
>
>
> Any point in adding data redundancy to joblib dumped objects?
>
Ah sorry for not being clear before, the steps in ipython were just to
demonstrate the compression issue not the flow of the code. I used an already
dumped object instead of retraining. [unless I
Any point in adding data redundancy to joblib dumped objects?
On Tue, Jan 22, 2013 at 5:35 PM, Gael Varoquaux <
gael.varoqu...@normalesup.org> wrote:
> On Tue, Jan 22, 2013 at 10:30:01PM +, Ark wrote:
> > /home/n7/newenv/lib/python2.6/site-
> > packages/sklearn/externals/joblib/numpy_pickle.
> Yes, documentation standard is probably a good idea. I would also push
> for API compatibility. And that brings me to the point raised in this
> thread: I really have to start working on proposing a good spec for the
> basic API of the scikit-learn. It's on my TODO list, somewhere not too
> far
Hi folks,
I think that the notion of affiliated or related projects is a great one
for many reasons.
On Tue, Jan 22, 2013 at 10:53:16AM -0800, Jake Vanderplas wrote:
> They have a well-defined set of criteria for packages seeking
> affiliation.
That's probably a good idea. We'll need to settle
On Tue, Jan 22, 2013 at 10:30:01PM +, Ark wrote:
> /home/n7/newenv/lib/python2.6/site-
> packages/sklearn/externals/joblib/numpy_pickle.pyc in read_zfile(file_handle)
> 69 assert len(data) == length, (
> 70 "Incorrect data length while decompressing %s."
> ---> 71
Hello,
I had been trying to dump a compressed joblib file (which was working fine
about a month ago). Previously I had an issue with amount of memory that
joblib compression took and it seemed that zlib was the issue. But I got more
memory to satisfy the problem.
However when I tried it
On Tue, Jan 22, 2013 at 4:14 PM, Lars Buitinck wrote:
> 2013/1/22 Andreas Mueller :
> > I think a system like that will be a good way to go forward.
>
> Me too -- I've been thinking about something like this for a
> scikit-sequence-learn project. Not that I have the time to start one,
> though :)
2013/1/22 Andreas Mueller :
> I think a system like that will be a good way to go forward.
Me too -- I've been thinking about something like this for a
scikit-sequence-learn project. Not that I have the time to start one,
though :)
> If we can't say what our estimators look like, it's very hard t
how about all related projects must guarantee a .fit method? The generality
of sklearn is why I use it personally.
On Tue, Jan 22, 2013 at 2:02 PM, Andreas Mueller
wrote:
> Hi Jake.
> Thanks for your input.
> I think a system like that will be a good way to go forward.
> For that to be possible,
Hi Jake.
Thanks for your input.
I think a system like that will be a good way to go forward.
For that to be possible, we first have to actually define the sklearn
interface, something
we postponed for quite a while now.
If we can't say what our estimators look like, it's very hard to make
require
+1 I like the idea of code promoted upstream for usefulness, I think it
will reduce bloat that typically comes with associated projects.
On Tue, Jan 22, 2013 at 1:53 PM, Jake Vanderplas <
vanderp...@astro.washington.edu> wrote:
> Hi all,
> I've been having some discussions with folks about the r
Hi all,
I've been having some discussions with folks about the relationship
between astropy and astroML, two complementary python astronomy
packages, and I wonder if some of these ideas could be useful as we
think about the future of scikit-learn.
In particular, astropy has the concept of "affi
Thanks everybody for the kind words :)
I'm glad I'm part of such a great community.
And people are already getting busy again on github!
Amazing
--
Master Visual Studio, SharePoint, SQL, ASP.NET, C# 2012, HTML5, CSS,
Hi Jake -
ok, this answers the question. I would say the standard definition of Euclidean
or Minkowski distance [i.e. with the root] is what we expect here, and also is
needed to satisfy the triangle inequality, so no problem.
Thanks, Tom
--
-
Institu
Am 22.01.2013 14:44, schrieb Yaroslav Halchenko:
> Let's push it push it push it (the 0.13 tag)!
>
I didn't even push the branch? It seems I was really tired yesterday.
I'll do it in ~5h when I'm at home.
--
Master Visual
Am 22.01.2013 14:44, schrieb Yaroslav Halchenko:
> Let's push it push it push it (the 0.13 tag)!
Damn, forgot to push the tags. Always forget that they are not pushed by
default.
Give me a second.
--
Master Visual Studio,
Let's push it push it push it (the 0.13 tag)!
On Tue, 22 Jan 2013, Robert Layton wrote:
>Congrats to all!
>On 22 January 2013 10:02, Andreas Mueller <[1]amuel...@ais.uni-bonn.de>
>wrote:
> Hi all.
> I am very happy to announce the release of scikit-learn 0.13.
> New f
Welldone everyone.. Thanks especially to Andy and managing the release so
well
regards
2013/1/22 Philipp Singer
> Great work as always guys!
>
> Eager to try out the new features, especially the feature hashing.
>
> Am 22.01.2013 00:02, schrieb Andreas Mueller:
> > Hi all.
> > I am very happy
Great work as always guys!
Eager to try out the new features, especially the feature hashing.
Am 22.01.2013 00:02, schrieb Andreas Mueller:
> Hi all.
> I am very happy to announce the release of scikit-learn 0.13.
> New features in this release include feature hashing for text processing,
> passi
Well done !!!
Arnaud
Le 22/01/2013 08:26, Gilles Louppe a écrit :
> Great job to all of you :)
>
> Gilles
>
> On 22 January 2013 07:57, Peter Prettenhofer
> wrote:
>> Great work guys - especially Andy - thanks a lot for making this happen!
>>
>> best,
>> Peter
>>
>> 2013/1/22 Gael Varoquaux :
great jobs folks ! As one could say "really really nice... [1]"
Cheers,
Alex
[1] http://www.youtube.com/watch?v=Nx7v815bYUw
--
Master Visual Studio, SharePoint, SQL, ASP.NET, C# 2012, HTML5, CSS,
MVC, Windows 8 Apps, Jav
Fixed the check button.
That was stupid :-/
I was so looking forward to viewing the answers this morning.
--
Master Visual Studio, SharePoint, SQL, ASP.NET, C# 2012, HTML5, CSS,
MVC, Windows 8 Apps, JavaScript and muc
On 01/22/2013 09:04 AM, Peter Prettenhofer wrote:
> according to the help the error msg show up "when the form creator
> stopped collecting responses by unchecking Accepting responses in the
> Form menu (under the Tools menu)." [1]
>
> [1] http://support.google.com/drive/bin/answer.py?hl=en&answer=
according to the help the error msg show up "when the form creator
stopped collecting responses by unchecking Accepting responses in the
Form menu (under the Tools menu)." [1]
[1] http://support.google.com/drive/bin/answer.py?hl=en&answer=1715669
2013/1/22 Mathieu Blondel :
> The link to the surv
38 matches
Mail list logo
