QST: How does pandas ensure backward compatibility with numpy and other deps

[maxkoretskyi]

Research

• I have searched the [pandas] tag on StackOverflow for similar questions.

• I have asked my usage related question on StackOverflow.

Link to question on StackOverflow

https://stackoverflow.com/questions/78808646

Question about pandas

If I check pandas dependencies requirements, I see the following:

+-------------------+--------------------------+
|     Package       | Minimum supported version|
+-------------------+--------------------------+
| NumPy             | 1.22.4                   |
+-------------------+--------------------------+
| python-dateutil   | 2.8.2                    |
+-------------------+--------------------------+
| pytz              | 2020.1                   |
+-------------------+--------------------------+
| tzdata            | 2022.7                   |
+-------------------+--------------------------+

I'm wondering how does pandas ensure that they are backward compatible with all the versions that were released after those specified in the table? I'm sure there are lots of tests, but it's not feasible to run tests against all possible permutations of those libs.

[mroeschke]

I'm wondering how does pandas ensure that they are backward compatible with all the versions that were released after those specified in the table?

There is no official CI testing with all the intermediate versions of these dependencies since as you mentioned it's not feasible. pandas only tests with these dependencies pinned to the minimum versions and unpinned (presumably the latest versions of these libraries).