Re: [Scikit-learn-general] What caused the HMM test failure

2012-10-29 Thread David Warde-Farley
On Mon, Oct 29, 2012 at 2:03 PM, Andreas Mueller wrote: > Hi everybody. > Olivier asked me to explain my commit: > https://github.com/scikit-learn/scikit-learn/commit/00d1055bf9f00e7024f9c4757262ffb852a61267 > On the ML, so here goes: > The code used > > as_float_array(startprob, copy=False) > > w

Re: [Scikit-learn-general] Dynamic Time Warping measure (DTW)

2012-10-29 Thread Stéfan van der Walt
On Mon, Oct 29, 2012 at 11:10 AM, Alexandre Gramfort wrote: > DTW in O(n^2) is super easy to implement. A good exercise with cython ! > Some implementations exist in O(n) although they give only good > approximations. Also see: https://github.com/scikit-image/scikit-image/blob/master/skimage/gr

Re: [Scikit-learn-general] Dynamic Time Warping measure (DTW)

2012-10-29 Thread didier vila
All, Someone is interested to implement it in the same time of me ? I can try to implement it , even my python is medium and my cython is void ! Where I should start ? my gibhub is pepgardiola. Didier > From: alexandre.gramf...@inria.fr > Date: Mon, 29 Oct 2012 19:10:29 +0100 > To: scikit-learn-g

Re: [Scikit-learn-general] Dynamic Time Warping measure (DTW)

2012-10-29 Thread Alexandre Gramfort
hi, DTW in O(n^2) is super easy to implement. A good exercise with cython ! Some implementations exist in O(n) although they give only good approximations. Alex On Mon, Oct 29, 2012 at 6:15 PM, Didier Vila wrote: > Olivier, > > it will be fantastic if the DTW can be implemented and be used for

[Scikit-learn-general] What caused the HMM test failure

2012-10-29 Thread Andreas Mueller
Hi everybody. Olivier asked me to explain my commit: https://github.com/scikit-learn/scikit-learn/commit/00d1055bf9f00e7024f9c4757262ffb852a61267 On the ML, so here goes: The code used as_float_array(startprob, copy=False) which I replaced with np.array(startprob, dtype=np.float) Basically as_

Re: [Scikit-learn-general] Dynamic Time Warping measure (DTW)

2012-10-29 Thread Didier Vila
Olivier, it will be fantastic if the DTW can be implemented and be used for any predictor of scikit learn. This implementation will leverage scikit learn and some of its current components for a time series analysis task. Regards. Didier Didier Vila, PhD | Risk -Original Message

Re: [Scikit-learn-general] Dynamic Time Warping measure (DTW)

2012-10-29 Thread Didier Vila
Leon, Thanks for your email. I can not install Rpy2 so I will go for mlpy. Thanks for your email and link. Didier Didier Vila, PhD | Risk | CapQuest Group Ltd | Fleet 27 | Rye Close | Fleet | Hampshire | GU51 2QQ | Fax: 0871 574 2992 | Email: dv...@capquestco.com

Re: [Scikit-learn-general] Dynamic Time Warping measure (DTW)

2012-10-29 Thread Leon Palafox
Didier Try using the DTW library in R with a wrapper to python, that seems to work ok. http://stackoverflow.com/questions/5695388/dynamic-time-warping-in-python On Mon, Oct 29, 2012 at 7:09 AM, Didier Vila wrote: > Olivier, > > Thanks > > Didier > > Didier Vila, PhD | Risk | CapQuest Group Ltd

Re: [Scikit-learn-general] List of Parameters and Attributes for Scikit-learn estimators

2012-10-29 Thread Jaques Grobler
Thanks for the feedback Satraijit, I'll make an optional ad-don once I've tidied the mess :) Kind Regards 2012/10/29 Satrajit Ghosh > hey jaques, > > this is really great! one useful thing to add, if possible, would be to > scrape the description of the parameter from the docstring. > > aside:

Re: [Scikit-learn-general] List of Parameters and Attributes for Scikit-learn estimators

2012-10-29 Thread Satrajit Ghosh
hey jaques, this is really great! one useful thing to add, if possible, would be to scrape the description of the parameter from the docstring. aside: if at some point in time, sklearn starts caring about provenance [1], this is a really good start at simply collecting terms. cheers, satra [1]

Re: [Scikit-learn-general] Dynamic Time Warping measure (DTW)

2012-10-29 Thread Didier Vila
Olivier, Thanks Didier Didier Vila, PhD | Risk | CapQuest Group Ltd | Fleet 27 | Rye Close | Fleet | Hampshire | GU51 2QQ | Fax: 0871 574 2992 | Email: dv...@capquestco.com -Original Message- From: Olivier Grisel [mailto:olivier.gri...@ensta.org] Sent: 29 October 2012 14:08 To: s

Re: [Scikit-learn-general] Dynamic Time Warping measure (DTW)

2012-10-29 Thread Olivier Grisel
2012/10/29 Didier Vila : > Good Afternoon all, > > Does Scikit learn have the dynamic time warping measure embedded in the > framework ? No. > If yes, can I use it with clustering/ svm tools ? > > If not, do you have any other alternatives / suggestions ? According to the wikipedia article: http

[Scikit-learn-general] Dynamic Time Warping measure (DTW)

2012-10-29 Thread Didier Vila
Good Afternoon all, Does Scikit learn have the dynamic time warping measure embedded in the framework ? If yes, can I use it with clustering/ svm tools ? If not, do you have any other alternatives / suggestions ? Regards Didier This e-mail is intended solely for the addressee, is strictly conf

Re: [Scikit-learn-general] List of Parameters and Attributes for Scikit-learn estimators

2012-10-29 Thread Jaques Grobler
Haha yeap...doing that now.. just thought I'd let the feedback come so long :) 2012/10/29 Andreas Mueller > On 10/29/2012 01:26 PM, Jaques Grobler wrote: > > Hi everyone. > > > > I've put up a gist with a list of all the parameters, how many times > > and where they occur, as well as for all the

Re: [Scikit-learn-general] List of Parameters and Attributes for Scikit-learn estimators

2012-10-29 Thread Andreas Mueller
On 10/29/2012 01:26 PM, Jaques Grobler wrote: > Hi everyone. > > I've put up a gist with a list of all the parameters, how many times > and where they occur, as well as for all the attributes. > The script that generates it is also in the gist - just know it's not > the prettiest due to borrowing

[Scikit-learn-general] List of Parameters and Attributes for Scikit-learn estimators

2012-10-29 Thread Jaques Grobler
Hi everyone. I've put up a gist with a list of all the parameters, how many times and where they occur, as well as for all the attributes. The script that generates it is also in the gist - just know it's not the prettiest due to borrowing data-gen code from tests and debugging things- I'm still m

Re: [Scikit-learn-general] I messed up or master severely broken

2012-10-29 Thread Andreas Mueller
On 10/29/2012 10:25 AM, Olivier Grisel wrote: > 2012/10/28 Gael Varoquaux : >> On Sun, Oct 28, 2012 at 09:38:31PM +, Andreas Mueller wrote: >>> Thanks for the quick fix :) >> Can you now build master? >> >> It works for me (and it's time for me to go to bed :} ). >> >> Gaël >> >> PS: looks like

Re: [Scikit-learn-general] I messed up or master severely broken

2012-10-29 Thread Olivier Grisel
2012/10/28 Gael Varoquaux : > On Sun, Oct 28, 2012 at 09:38:31PM +, Andreas Mueller wrote: >> Thanks for the quick fix :) > > Can you now build master? > > It works for me (and it's time for me to go to bed :} ). > > Gaël > > PS: looks like we need better/more buildbots We only test on linux 6