Mailman 3 - Pandas-dev

ANN: pandas 3.0
by Joris Van den Bossche Jan. 22, 2026

Jan. 22, 2026

We are very pleased to announce the release of pandas 3.0.0, a major release from the pandas 2.x series. This release includes various new features, bug fixes, and performance improvements, as well as possible breaking changes. Highlights include: - Dedicated string data type by default <https://pandas.pydata.org/docs/whatsnew/v3.0.0.html#whatsnew-300-enhancemen…> - Consistent copy/view behaviour with Copy-on-Write <https://pandas.pydata.org/docs/whatsnew/v3.0.0.html#whatsnew-300-enhancemen…> (CoW) (a.k.a. getting rid of the SettingWithCopyWarning) - New default resolution for datetime-like data <https://pandas.pydata.org/docs/whatsnew/v3.0.0.html#whatsnew-300-api-breaki…> - Initial support for the new pd.col syntax <https://pandas.pydata.org/docs/whatsnew/v3.0.0.html#whatsnew-300-enhancemen…> See the announcement blog post <https://pandas.pydata.org/community/blog/pandas-3.0.html> and the detailed release notes <https://pandas.pydata.org/docs/whatsnew/v3.0.0.html> for a list of all the changes. The pandas 3.0 release removed a functionality that was deprecated in previous releases. It is recommended to first upgrade to pandas 2.3 and to ensure your code is working without warnings, before upgrading to pandas 3.0. In addition, as a major release, pandas 3.0 includes some breaking changes that may require updates to your code. An overview of all potentially breaking changes can be found in the Backwards incompatible API changes <https://pandas.pydata.org/docs/whatsnew/v3.0.0.html#backwards-incompatible-…> section of the release notes. Pandas 3.0.0 supports Python 3.11 and higher. The release can be installed from PyPI: python -m pip install --upgrade pandas==3.0.* or from conda-forge: conda install -c conda-forge pandas=3.0 Please report any issues with the release on the pandas issue tracker <https://github.com/pandas-dev/pandas/issues/new/choose>. Thanks to all the contributors who made this release possible!

1 0

Next pandas community meeting - Wednesday January 22, UTC 18:00
by Joris Van den Bossche Jan. 14, 2026

Jan. 14, 2026

Hi all, A reminder that tomorrow we have the next bi-monthly (twice a month) dev call (Wednesday, January 22) at 18:00 UTC. Our calendar is at https://pandas.pydata.org/docs/dev/development/community.html to check your local time. The pandas Community Meeting is a regular sync meeting for the project's maintainers which is open to the community. All are welcome to attend! Video Call: https://numfocus-org.zoom.us/j/84024486573?pwd=QnVOUkozc3VzeXVsYnJEM1h4QVBG… Meeting notes: https://docs.google.com/document/u/1/d/1tGbTiYORHiSPgVMXawiweGJlBw5dOkVJLY-… Joris

1 2

ANN: pandas 3.0.0 release candidate ready for testing!
by Joris Van den Bossche Dec. 10, 2025

Dec. 10, 2025

We are happy to announce the first *release candidate* of pandas 3.0.0, released last week. You can find the the list of changes in 3.0.0 in the release notes page <https://pandas.pydata.org/docs/dev/whatsnew/v3.0.0.html>. Users having pandas code in production and maintainers of libraries with pandas as a dependency are *strongly* recommended to run their test suites with the release candidate, and report any breaking change to our issue tracker <https://github.com/pandas-dev/pandas/issues/> before the official 3.0.0 release. Especially the changes around the new default string dtype and Copy-on-Write are expected to require some code changes. See the linked release notes for details how to test and update your code. The release candidate is available on PyPI and conda-forge, for example: python -m pip install --upgrade pandas==3.0.0rc0 or conda install -c conda-forge/label/pandas_rc pandas==3.0.0rc0

1 0

Introduction and Request for Guidance for GSoC Contributions
by M noman Shafi Nov. 30, 2025

Nov. 30, 2025

Hello, I hope you are doing well. My name is Muhammad, and I am preparing to contribute under the Python Software Foundation for GSoC. I am currently focusing on the Pandas project and have begun exploring issues, setting up the development environment, and learning the workflow. I wanted to introduce myself to the community and request general guidance on how I can effectively contribute. If there are recommended channels, documentation, beginner-friendly issues, or contribution practices that I should follow, I would really appreciate any direction. I am committed to learning, improving, and contributing meaningfully to the project. Any advice or suggestions from the mentors or experienced contributors would be extremely valuable. Thank you for your time. Best regards, Muhammad

1 0

ANN: pandas 2.3.2 - with preview of upcoming pandas 3.0 features
by Joris Van den Bossche Aug. 22, 2025

Aug. 22, 2025

Hi all, We are pleased to announce the release of pandas v2.3.2. See the release notes <https://pandas.pydata.org/docs/whatsnew/v2.3.2.html> for a list of all the changes. And as usual, the latest release can be installed from PyPI or conda-forge. While this patch release includes some bug fixes, we mostly want to draw attention to the opt-in features included in pandas 2.3, as a preview for changes coming in pandas 3.0. Starting with pandas 3.0, the dedicated string data type and the Copy-on-Write mode will be enabled by default. See Upcoming changes in pandas 3.0 <https://pandas.pydata.org/docs/whatsnew/v2.3.0.html#whatsnew-230-upcoming-c…> for more details. Summarizing here: - *String dtype*: Historically, pandas has represented string columns with NumPy's "object" data type. With pandas 3.0, the dedicated string data type ("str") will be inferred by default. This gives a much clearer user experience, and, if pyarrow is installed, will also give a performance boost for string operations. - *Copy-on-Write*: this is the technical term for the underlying implementation, but the main behaviour change is that the result of any indexing operation or any method returning a new DataFrame or Series, will always *behave as if* it were a copy. Thus, no longer modifying a DataFrame through chaining, and no longer a SettingWithCopyWarning! This are big changes, and changes that most likely will have an impact on your pandas code! Therefore, it would be great if you can help test those changes, and to make that easier they are included in pandas 2.3 behind a flag. You can enable both with: import pandas as pd pd.options.future.infer_string = True pd.options.mode.copy_on_write = True Check the string dtype migration guide <https://pandas.pydata.org/docs/user_guide/migration-3-strings.html> and Copy-on-Write user guide <https://pandas.pydata.org/docs/user_guide/copy_on_write.html> for more details about how to test, prepare and update your code to be compatible with those changes. Please report any issues you run into on the pandas issue tracker <https://github.com/pandas-dev/pandas/issues/new/choose>!

1 0

ANN: pandas 2.3.0
by emailformattr＠gmail.com June 8, 2025

June 8, 2025

We are pleased to announce the release of pandas v2.3.0. This is a minor release in the 2.3.x series. We recommend that all users in the 2.3.x series upgrade to this version. See the release notes (https://pandas.pydata.org/pandas-docs/version/2.3.0/whatsnew/v2.3.0.html) for a list of all the changes. The release can be installed from PyPI python -m pip install --upgrade pandas==2.3.0 Or from conda-forge conda install -c conda-forge pandas==2.3.0 Please report any issues with the release on the pandas issue tracker (https://github.com/pandas-dev/pandas/issues/new/choose). Thanks to all the contributors (https://pandas.pydata.org/pandas-docs/version/2.3.0/whatsnew/v2.3.0.html#co…) who made this release possible.

1 0

Default read_csv engine in pandas 3.0
by Marc Garcia June 4, 2025

June 4, 2025

Hi there, I was thinking the upcoming pandas 3.0 release would be a good opportunity to change the default engine for pandas.read_csv, and make it be pyarrow for users who have pyarrow installed. It could work in a similar way as read_parquet that the default depends on what's installed (with the difference that the C engine will always be installed). The motivation for the change is that I think in general (at least in my experience) the pyarrow engine runs in half the time as the C engine. And I think the behavior will be somehow consistent to what's been discussed for strings, so I think it'll be easy for users to understand and get used to. Since read_csv is such a common function it'd be good to get feedback from people who thinks this is not a good idea, and understand the problems this can cause. For users with pyarrow not installed in the environment there wouldn't be any change in how pandas works. Thanks!

4 6

Request for guidance on issues for upcoming PyData Yerevan pandas sprint
by Suren Poghosyan April 24, 2025

April 24, 2025

Dear Pandas Team, I am Suren Poghosyan, an Organizational Committee Member at PyData Yerevan. As you may already know, last summer we hosted an open-source contribution sprint focused on *pandas* library, in collaboration with Patrick Höfler, at the American University of Armenia. We are currently planning to run a follow-up sprint independently and would appreciate your guidance on which issues in the *pandas* GitHub repository are the most appropriate for a 2-3 hour contribution session. Furthermore, feel free to share any relevant issue which you would like to proceed with, despite our previous specification and time span. On top of that, we are reaching out to make sure we’re following the proper contribution guidelines and not creating unnecessary noise or inconveniences in the issue tracker. We aim to contribute meaningfully and respectfully. For your reference, here are the articles about our previous sprint: - AUA to Host Inaugural PyData Yerevan Open Source pandas Sprint <https://newsroom.aua.am/2024/06/10/aua-to-host-inaugural-pydata-yerevan-ope…> - PyData Yerevan Open Source pandas Sprint <https://newsroom.aua.am/event/pydata-yerevan-open-source-pandas-sprint/> - PyData Yerevan hosted the inaugural Open Source pandas Sprint with Patrick Höfler - Linkedin <https://www.linkedin.com/posts/pydata-yerevan_yesterday-pydata-yerevan-host…> We are looking forward to contributing again and strengthening our local culture of open-source collaboration. Best regards, Suren Poghosyan Organizational Committee Member PyData Yerevan

1 0

Next pandas community meeting - Wednesday October 9, UTC 18:00
by Joris Van den Bossche Nov. 12, 2024

Nov. 12, 2024

Hi all, A reminder that the next bi-monthly (twice a month) dev call is later today (Wednesday, October 9) at 18:00 UTC. Our calendar is at https://pandas.pydata.org/docs/dev/development/community.html to check your local time. The pandas Community Meeting is a regular sync meeting for the project's maintainers which is open to the community. All are welcome to attend! Video Call: https://numfocus-org.zoom.us/j/84024486573?pwd=QnVOUkozc3VzeXVsYnJEM1h4QVBG… Meeting notes: https://docs.google.com/document/u/1/d/1tGbTiYORHiSPgVMXawiweGJlBw5dOkVJLY-… Joris

2 2

TSV Files Not Supported In Latest Version
by Aliyan Anwar Oct. 13, 2024

Oct. 13, 2024

hello greetings from aliyan ! am facing this error when opening a tsv file in latest version, and another issue is . import pandas as pd # Reading a TSV file df = pd.read_csv('file_name.tsv', sep='\t') 1) I am trying to run pandas but getting this error ModuleNotFoundError: No module named 'numpy.rec' _______________________________________________________ 2) When opening a pickle file of type list -> of pandas df. with open('day_features_data_lagby_14_v2.bin', 'rb') as f: currency_ohlc = pickle.load((f)) TypeError: Argument 'placement' has incorrect type (expected pandas._libs.internals.BlockPlacement, got slice)

1 1