From dabruro at gmail.com Wed May 4 13:37:52 2022 From: dabruro at gmail.com (David R) Date: Wed, 4 May 2022 13:37:52 -0400 Subject: [scikit-learn] suggestion for metrics.classification_report Message-ID: [Is there a better place to make a suggestion like this, other than this mailing list?] classification_report redundantly displays the *accuracy *on its own line even though it is already also showing the *weighted average recall* which is just another name for the *accuracy*! Instead of this: precision recall f1-score support class 0 0.50 1.00 0.67 1 class 1 0.00 0.00 0.00 1 class 2 1.00 0.67 0.80 3 *accuracy * *0.60 5* macro avg 0.50 0.56 0.49 5weighted avg 0.70 0.60 0.61 5 How about something more like the following to avoid the redundant accuracy line above (and perhaps to *also *label *balanced *accuracy while we're at it): precision recall f1-score support class 0 0.50 1.00 0.67 1 class 1 0.00 0.00 0.00 1 class 2 1.00 0.67 0.80 3 macro avg 0.50 0.56 * 0.49 5 * balanced accuracyweighted avg 0.70 0.60 ** 0.61 5 ** accuracy David Rosen -------------- next part -------------- An HTML attachment was scrubbed... URL: From thomasjpfan at gmail.com Sat May 7 18:19:43 2022 From: thomasjpfan at gmail.com (Thomas J. Fan) Date: Sat, 7 May 2022 18:19:43 -0400 Subject: [scikit-learn] scikit-learn monthly developer meeting: Monday May 30, 2022 Message-ID: Dear all, The scikit-learn developer monthly meeting will take place on Monday May 30, 2022 at 13:00 UTC. - Video call link: https://meet.google.com/ews-uszu-djs - Meeting notes / agenda: https://hackmd.io/0yokz72CTZSny8y3Re648Q - Local times: https://www.timeanddate.com/worldclock/meetingdetails.html?year=2022&month=5&day=30&hour=13&min=0&sec=0&p1=1440&p2=240&p3=248&p4=195&p5=179&p6=224 The goal of this meeting is to discuss ongoing development topics for the project. Everybody is welcome. As usual, please follow the code of conduct of the project: https://github.com/scikit-learn/scikit-learn/blob/main/CODE_OF_CONDUCT.md Regards, Thomas -------------- next part -------------- An HTML attachment was scrubbed... URL: From jeremie.du-boisberranger at inria.fr Thu May 12 11:20:24 2022 From: jeremie.du-boisberranger at inria.fr (Jeremie du Boisberranger) Date: Thu, 12 May 2022 17:20:24 +0200 Subject: [scikit-learn] [ANN] scikit-learn 1.1 release In-Reply-To: References: Message-ID: Hi everyone, We're happy to announce the 1.1 release which you can install via pip or conda: ??? pip install -U scikit-learn or (soon) ??? conda install -c conda-forge scikit-learn The wheels for arm64 are not available yet on PyPI though. We'll add them as soon as possible. You can read the release highlights under https://scikit-learn.org/stable/auto_examples/release_highlights/plot_release_highlights_1_1_0.html and the long list of the changes under https://scikit-learn.org/stable/whats_new/v1.1.html This version supports Python versions 3.8 to 3.10. A big thanks to all contributors who helped on this release. Regards, J?r?mie, On the behalf of the scikit-learn maintainer team. From adrin.jalali at gmail.com Thu May 12 11:39:43 2022 From: adrin.jalali at gmail.com (Adrin) Date: Thu, 12 May 2022 17:39:43 +0200 Subject: [scikit-learn] [ANN] scikit-learn 1.1 release In-Reply-To: References: Message-ID: Thanks a ton for the efforts Jeremie! Delighted to see the release out! On Thu, May 12, 2022 at 5:21 PM Jeremie du Boisberranger < jeremie.du-boisberranger at inria.fr> wrote: > Hi everyone, > > We're happy to announce the 1.1 release which you can install via pip or > conda: > > pip install -U scikit-learn > > or (soon) > > conda install -c conda-forge scikit-learn > > The wheels for arm64 are not available yet on PyPI though. We'll add > them as soon as possible. > > You can read the release highlights under > > https://scikit-learn.org/stable/auto_examples/release_highlights/plot_release_highlights_1_1_0.html > and the long list of the changes under > https://scikit-learn.org/stable/whats_new/v1.1.html > > This version supports Python versions 3.8 to 3.10. > > A big thanks to all contributors who helped on this release. > > Regards, > J?r?mie, > On the behalf of the scikit-learn maintainer team. > _______________________________________________ > scikit-learn mailing list > scikit-learn at python.org > https://mail.python.org/mailman/listinfo/scikit-learn > -------------- next part -------------- An HTML attachment was scrubbed... URL: From bertrand.thirion at inria.fr Thu May 12 12:09:32 2022 From: bertrand.thirion at inria.fr (bthirion) Date: Thu, 12 May 2022 18:09:32 +0200 Subject: [scikit-learn] [ANN] scikit-learn 1.1 release In-Reply-To: References: Message-ID: <72d32b5e-b85b-dc24-f0ba-11755f0b88d4@inria.fr> Congratulations ! B Le 12/05/2022 ? 17:20, Jeremie du Boisberranger a ?crit?: > Hi everyone, > > We're happy to announce the 1.1 release which you can install via pip > or conda: > > ??? pip install -U scikit-learn > > or (soon) > > ??? conda install -c conda-forge scikit-learn > > The wheels for arm64 are not available yet on PyPI though. We'll add > them as soon as possible. > > You can read the release highlights under > https://scikit-learn.org/stable/auto_examples/release_highlights/plot_release_highlights_1_1_0.html > and the long list of the changes under > https://scikit-learn.org/stable/whats_new/v1.1.html > > This version supports Python versions 3.8 to 3.10. > > A big thanks to all contributors who helped on this release. > > Regards, > J?r?mie, > On the behalf of the scikit-learn maintainer team. > _______________________________________________ > scikit-learn mailing list > scikit-learn at python.org > https://mail.python.org/mailman/listinfo/scikit-learn From olivier.grisel at ensta.org Thu May 12 12:08:32 2022 From: olivier.grisel at ensta.org (Olivier Grisel) Date: Thu, 12 May 2022 18:08:32 +0200 Subject: [scikit-learn] [ANN] scikit-learn 1.1 release In-Reply-To: References: Message-ID: Congrats Jeremie and everybody who contributed to this release! This is a great achievement. -- Olivier From gael.varoquaux at normalesup.org Thu May 12 12:43:14 2022 From: gael.varoquaux at normalesup.org (Gael Varoquaux) Date: Thu, 12 May 2022 18:43:14 +0200 Subject: [scikit-learn] [ANN] scikit-learn 1.1 release In-Reply-To: References: Message-ID: <20220512164314.4m5wgyklf6tugvem@phare.normalesup.org> Wohoo!! Thank you so much. This is so exciting: all those nice improvements reaching so much users. Ga?l On Thu, May 12, 2022 at 05:20:24PM +0200, Jeremie du Boisberranger wrote: > Hi everyone, > We're happy to announce the 1.1 release which you can install via pip or > conda: > ??? pip install -U scikit-learn > or (soon) > ??? conda install -c conda-forge scikit-learn > The wheels for arm64 are not available yet on PyPI though. We'll add them as > soon as possible. > You can read the release highlights under https://scikit-learn.org/stable/auto_examples/release_highlights/plot_release_highlights_1_1_0.html > and the long list of the changes under > https://scikit-learn.org/stable/whats_new/v1.1.html > This version supports Python versions 3.8 to 3.10. > A big thanks to all contributors who helped on this release. > Regards, > J?r?mie, > On the behalf of the scikit-learn maintainer team. > _______________________________________________ > scikit-learn mailing list > scikit-learn at python.org > https://mail.python.org/mailman/listinfo/scikit-learn -- Gael Varoquaux Research Director, INRIA http://gael-varoquaux.info http://twitter.com/GaelVaroquaux From alexandre.gramfort at inria.fr Thu May 12 15:25:06 2022 From: alexandre.gramfort at inria.fr (Alexandre Gramfort) Date: Thu, 12 May 2022 21:25:06 +0200 Subject: [scikit-learn] [ANN] scikit-learn 1.1 release In-Reply-To: <20220512164314.4m5wgyklf6tugvem@phare.normalesup.org> References: <20220512164314.4m5wgyklf6tugvem@phare.normalesup.org> Message-ID: awesome ! ? Alex On Thu, May 12, 2022 at 6:53 PM Gael Varoquaux < gael.varoquaux at normalesup.org> wrote: > Wohoo!! Thank you so much. This is so exciting: all those nice > improvements reaching so much users. > > Ga?l > > On Thu, May 12, 2022 at 05:20:24PM +0200, Jeremie du Boisberranger wrote: > > Hi everyone, > > > We're happy to announce the 1.1 release which you can install via pip or > > conda: > > > pip install -U scikit-learn > > > or (soon) > > > conda install -c conda-forge scikit-learn > > > The wheels for arm64 are not available yet on PyPI though. We'll add > them as > > soon as possible. > > > You can read the release highlights under > https://scikit-learn.org/stable/auto_examples/release_highlights/plot_release_highlights_1_1_0.html > > and the long list of the changes under > > https://scikit-learn.org/stable/whats_new/v1.1.html > > > This version supports Python versions 3.8 to 3.10. > > > A big thanks to all contributors who helped on this release. > > > Regards, > > J?r?mie, > > On the behalf of the scikit-learn maintainer team. > > _______________________________________________ > > scikit-learn mailing list > > scikit-learn at python.org > > https://mail.python.org/mailman/listinfo/scikit-learn > > -- > Gael Varoquaux > Research Director, INRIA > http://gael-varoquaux.info http://twitter.com/GaelVaroquaux > _______________________________________________ > scikit-learn mailing list > scikit-learn at python.org > https://mail.python.org/mailman/listinfo/scikit-learn > -------------- next part -------------- An HTML attachment was scrubbed... URL: From norbert at preining.info Thu May 12 19:22:11 2022 From: norbert at preining.info (Norbert Preining) Date: Fri, 13 May 2022 08:22:11 +0900 Subject: [scikit-learn] [ANN] scikit-learn 1.1 release In-Reply-To: References: Message-ID: > We're happy to announce the 1.1 release which you can install via pip or Congratulations to the team! Norbert -- PREINING Norbert https://www.preining.info Mercari Inc. + IFMGA Guide + TU Wien + TeX Live GPG: 0x860CDC13 fp: F7D8 A928 26E3 16A1 9FA0 ACF0 6CAC A448 860C DC13 From mahmood.nt at gmail.com Fri May 13 04:35:55 2022 From: mahmood.nt at gmail.com (Mahmood Naderan) Date: Fri, 13 May 2022 10:35:55 +0200 Subject: [scikit-learn] View full sized k_means.labels_ Message-ID: Hi, I have used the following lines of codes k_means = KMeans(n_clusters=i, random_state=4).fit(principalComponents_dataFrame) print(k_means.labels_) But the problem is for large vectors of labels, I see shortened version like this: [4 4 0 ... 0 0 0] How can I force it to print the full length vector? Regards, Mahmood From adrin.jalali at gmail.com Fri May 13 05:59:12 2022 From: adrin.jalali at gmail.com (Adrin) Date: Fri, 13 May 2022 11:59:12 +0200 Subject: [scikit-learn] View full sized k_means.labels_ In-Reply-To: References: Message-ID: Hi there, You can set print options for numpy: https://stackoverflow.com/questions/1987694/how-to-print-the-full-numpy-array-without-truncation But you most probably don't want to do that. If you need to print them, you can iterate through them and print them, or iterate and create a string that you'd like to print, for example: ", ".join([str(x) for x in k_means.labels_]) On Fri, May 13, 2022 at 10:37 AM Mahmood Naderan wrote: > Hi, > I have used the following lines of codes > > k_means = KMeans(n_clusters=i, > random_state=4).fit(principalComponents_dataFrame) > print(k_means.labels_) > > But the problem is for large vectors of labels, I see shortened > version like this: > > [4 4 0 ... 0 0 0] > > How can I force it to print the full length vector? > > Regards, > Mahmood > _______________________________________________ > scikit-learn mailing list > scikit-learn at python.org > https://mail.python.org/mailman/listinfo/scikit-learn > -------------- next part -------------- An HTML attachment was scrubbed... URL: From jmzhl at uchicago.edu Sun May 15 14:07:24 2022 From: jmzhl at uchicago.edu (Jonathan Li) Date: Sun, 15 May 2022 18:07:24 +0000 Subject: [scikit-learn] Multiplicative Weights Update Algorithm Contribution Message-ID: Dear the Scikit-learn community, I am a researcher at UChicago working on the Multiplicative Weights Update method, a widely used, effective meta-algorithm with broad applications that fits nicely with ensemble methods. The following well cited paper presents the algorithm neatly: http://www.theoryofcomputing.org/articles/v008a006/ I would like to propose adding this algorithm to Scikit-learn, which would be of great benefit to the machine learning research community. It would integrate nicely with the existing `VotingClassifier` and `VotingRegressor` algorithms. Please let me know if you need more information and what you think. I would be more than happy to take charge of this contribution given my experience with the algorithm and existing code built in Python. Best, Jonathan From g.lemaitre58 at gmail.com Mon May 16 15:04:52 2022 From: g.lemaitre58 at gmail.com (=?utf-8?Q?Guillaume_Lema=C3=AEtre?=) Date: Mon, 16 May 2022 21:04:52 +0200 Subject: [scikit-learn] ANN: imbalanced-learn 0.9.1 released Message-ID: Hi all, We are happy to announce the 0.9.1 version of imbalanced-learn. imbalanced-learn is a toolbox aiming at providing a wide range of methods to cope with the problem of imbalanced data sets frequently encountered in machine learning and data mining. This release makes sure to be fully compatible with scikit-learn 1.1.0 pip install -U imbalanced-learn The conda-forge builds will be available shortly, which you can then install using: conda install -c conda-forge imbalanced-learn For more information, examples, and documentation, please visit: https://imbalanced-learn.org -- Guillaume Lemaitre Scikit-learn @ Inria Foundation https://glemaitre.github.io/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From taldcroft at gmail.com Wed May 18 16:06:20 2022 From: taldcroft at gmail.com (Tom Aldcroft) Date: Wed, 18 May 2022 16:06:20 -0400 Subject: [scikit-learn] Experience with black formatting in scikit-learn for astropy Message-ID: Hi - The astropy core is currently considering implementing code formatting with black, much as scikit-learn did in 2020 ( https://github.com/scikit-learn/scikit-learn/pull/18948). Not surprisingly there are a wide range of opinions and concerns. My goal in writing is to request feedback from the scikit-learn community about how this transition has impacted development and community engagement in the time since the transition. Has there been specific feedback from contributors (both experienced and new) related to black autoformatting? - Ease of actually running black and meeting the black-format standard for PRs. Have there been contributor problems that required assistance? - Style considerations? - Technical problems like backports? Are there any regrets or things you would change? Do you have any advice for the astropy project? Any feedback from your experience would be most appreciated, especially coming from a more science-oriented project like scikit-learn (as opposed to other large projects like django or pytest that have also adopted black). Cheers, Tom (@taldcroft on GitHub) -------------- next part -------------- An HTML attachment was scrubbed... URL: From g.lemaitre58 at gmail.com Thu May 19 05:16:28 2022 From: g.lemaitre58 at gmail.com (=?UTF-8?Q?Guillaume_Lema=C3=AEtre?=) Date: Thu, 19 May 2022 11:16:28 +0200 Subject: [scikit-learn] Experience with black formatting in scikit-learn for astropy In-Reply-To: References: Message-ID: I just answer in the text below. This is my 2c. Hope this helps Cheers, On Wed, 18 May 2022 at 22:08, Tom Aldcroft wrote: > Hi - > > The astropy core is currently considering implementing code formatting > with black, much as scikit-learn did in 2020 ( > https://github.com/scikit-learn/scikit-learn/pull/18948). Not > surprisingly there are a wide range of opinions and concerns. > > My goal in writing is to request feedback from the scikit-learn community > about how this transition has impacted development and community engagement > in the time since the transition. > > Has there been specific feedback from contributors (both experienced and > new) related to black autoformatting? > > - Ease of actually running black and meeting the black-format standard > for PRs. Have there been contributor problems that required assistance? > > I think that this is the main drawback. One can argue that using `pre-commit` will do all the necessary changes but we still get contributors pushing without `pre-commit` being installed and stalled with an Azure CI. Since the CI could be quite overwhelming for new contributors, I could consider this part as an additional entrance barrier. > > - Style considerations? > > I assume that you get used to it even if anyone might get a little cranky on some styling at first. At least, it allows having any discussion regarding (some) style coding during PR reviews. I think that this is the main advantage for us. > > - Technical problems like backports? > > Not really. I think that it was quite smooth. We only get a commit in the history that is annoying while trying to blame with `git`. > Are there any regrets or things you would change? > Personally not ;). We might intend to use `isort` even. > Do you have any advice for the astropy project? > > Any feedback from your experience would be most appreciated, especially > coming from a more science-oriented project like scikit-learn (as opposed > to other large projects like django or pytest that have also adopted black). > > Cheers, > Tom (@taldcroft on GitHub) > > _______________________________________________ > scikit-learn mailing list > scikit-learn at python.org > https://mail.python.org/mailman/listinfo/scikit-learn > -- Guillaume Lemaitre Scikit-learn @ Inria Foundation https://glemaitre.github.io/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From olivier.grisel at ensta.org Thu May 19 07:50:36 2022 From: olivier.grisel at ensta.org (Olivier Grisel) Date: Thu, 19 May 2022 13:50:36 +0200 Subject: [scikit-learn] Experience with black formatting in scikit-learn for astropy In-Reply-To: References: Message-ID: I agree with Guillaume's answers. I think it was a net benefit, even though it might be a bit annoying to get the tooling right for first time contributors. We can probably improve this by making the error messages on the CI more directive on how to fix formatting issues by given copy-pastable commands to install and run black in your branch. Otherwise, I really like just pressing shift-ctrl-i to fix the formatting when editing code in VS Code. -- Olivier From taldcroft at gmail.com Thu May 19 09:02:23 2022 From: taldcroft at gmail.com (Tom Aldcroft) Date: Thu, 19 May 2022 09:02:23 -0400 Subject: [scikit-learn] Experience with black formatting in scikit-learn for astropy In-Reply-To: References: Message-ID: Thanks Guillaume and Olivier for the quick feedback. There are enough people on the fence in astropy that information like this is quite valuable. If anyone is interested, here is the monster thread. ;-) https://groups.google.com/g/astropy-dev/c/6cRJCMgaFyM Best, Tom On Thu, May 19, 2022 at 7:51 AM Olivier Grisel wrote: > I agree with Guillaume's answers. > > I think it was a net benefit, even though it might be a bit annoying > to get the tooling right for first time contributors. We can probably > improve this by making the error messages on the CI more directive on > how to fix formatting issues by given copy-pastable commands to > install and run black in your branch. > > Otherwise, I really like just pressing shift-ctrl-i to fix the > formatting when editing code in VS Code. > > -- > Olivier > _______________________________________________ > scikit-learn mailing list > scikit-learn at python.org > https://mail.python.org/mailman/listinfo/scikit-learn > -------------- next part -------------- An HTML attachment was scrubbed... URL: From g.lemaitre58 at gmail.com Thu May 19 09:42:54 2022 From: g.lemaitre58 at gmail.com (=?UTF-8?Q?Guillaume_Lema=C3=AEtre?=) Date: Thu, 19 May 2022 15:42:54 +0200 Subject: [scikit-learn] [ANN] scikit-learn 1.1.1 is online! Message-ID: scikit-learn 1.1.1 is out on pypi.org and conda-forge! This is a small maintenance release that fixes a couple of regressions: https://scikit-learn.org/dev/whats_new/v1.1.html#version-1-1-1 Notably, if you are using tree-based models (i.e. decision tree, random forest, gradient boosting), we corrected a computational performance regression introduced in 1.1.0. You can upgrade with pip as usual: pip install -U scikit-learn The conda-forge builds will be available shortly, which you can then install using: conda install -c conda-forge scikit-learn Thanks again to all the contributors. On behalf of the scikit-learn maintainer team. -- Guillaume Lemaitre Scikit-learn @ Inria Foundation https://glemaitre.github.io/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From olivier.grisel at ensta.org Thu May 19 10:36:35 2022 From: olivier.grisel at ensta.org (Olivier Grisel) Date: Thu, 19 May 2022 16:36:35 +0200 Subject: [scikit-learn] [ANN] scikit-learn 1.1.1 is online! In-Reply-To: References: Message-ID: Thank you to all the contributors who reported bugs, minimal reproducers and fixes, and thank you Guillaume for getting this bugfix release out so timely \o/ -- Olivier From olivier.grisel at ensta.org Thu May 19 10:37:12 2022 From: olivier.grisel at ensta.org (Olivier Grisel) Date: Thu, 19 May 2022 16:37:12 +0200 Subject: [scikit-learn] [ANN] scikit-learn 1.1.1 is online! In-Reply-To: References: Message-ID: BTW, this is now stable to the URL https://scikit-learn.org/stable/whats_new/v1.1.html#version-1-1-1 also works :) From g.lemaitre58 at gmail.com Wed May 25 17:16:36 2022 From: g.lemaitre58 at gmail.com (=?UTF-8?Q?Guillaume_Lema=C3=AEtre?=) Date: Wed, 25 May 2022 23:16:36 +0200 Subject: [scikit-learn] Announcement of EuroSciPy 2022 Message-ID: Dear all, The following conference could be of interest. Please find the CfP announcement of EuroSciPy 2022 below. --- We are happy to announce that EuroSciPy, the 14th European Conference on Python in Science, will be back as an in-person event. EuroSciPy 2022 will take place from Monday, August 29 to Friday, September 2 in Basel, Switzerland. The Call for Proposals will be open until Monday, 6 June Here's your chance to present a talk, a tutorial or a poster and share your thoughts at EuroSciPy 2022: https://pretalx.com/euroscipy-2022/cfp About EuroSciPy 2022 The EuroSciPy meeting is a cross-disciplinary gathering focused on the use and development of the Python language in scientific research. This event strives to bring together both users and developers of scientific tools, as well as academic research and state-of-the-art industry. Conference layout: - Mon-Tue, Aug 29-30: Tutorials. - Wed-Thu, Aug 31-Sep 1: Talks. - Fri, Sep 2: Sprints. Tickets Ticket sales will open beginning of June. This year we will offer both in-person and remote-tickets., having limited in-person attendance to 300 people in total. The remaining tickets are reserved for speakers, sponsors and volunteers. Any questions? Check out our website . All the best, -- Guillaume Lemaitre Scikit-learn @ Inria Foundation https://glemaitre.github.io/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From solegalli at protonmail.com Sun May 29 03:40:04 2022 From: solegalli at protonmail.com (Sole Galli) Date: Sun, 29 May 2022 07:40:04 +0000 Subject: [scikit-learn] View full sized k_means.labels_ In-Reply-To: References: Message-ID: Maybe with numpy.set_printoptions? See thread here: https://stackoverflow.com/questions/1987694/how-to-print-the-full-numpy-array-without-truncation Soledad Galli https://www.trainindata.com/ Sent with Proton Mail secure email. ------- Original Message ------- On Friday, May 13th, 2022 at 10:35 AM, Mahmood Naderan wrote: > Hi, > I have used the following lines of codes > > k_means = KMeans(n_clusters=i, > random_state=4).fit(principalComponents_dataFrame) > print(k_means.labels_) > > But the problem is for large vectors of labels, I see shortened > version like this: > > [4 4 0 ... 0 0 0] > > How can I force it to print the full length vector? > > Regards, > Mahmood > _______________________________________________ > scikit-learn mailing list > scikit-learn at python.org > https://mail.python.org/mailman/listinfo/scikit-learn From seralouk at hotmail.com Tue May 31 04:59:41 2022 From: seralouk at hotmail.com (Serafeim Loukas) Date: Tue, 31 May 2022 08:59:41 +0000 Subject: [scikit-learn] Kernels in logistic regression model Message-ID: <4FEE3001-AF6B-48B0-ACCA-5FD9225C0D4C@hotmail.com> Dear sklearn community, More and more people seem to be interested in using kernels in the logistic regression models. Initially, I saw this post here, and I proposed a solution for the ANOVA and Nystr?m kernel methods. Would it make sense to add these in the next sklearn version? Best, Makis -------------- next part -------------- An HTML attachment was scrubbed... URL: