Rigid requirement version of pandas pandas==1.5.3 in requirements.txt hinders innovation

[mariakesa]

Describe the use case that is addressed by this feature.
I am developing a library for analyzing Allen data, and I want to integrate AI agents with LLM's into the library. However, I can't align dependencies, specifically a hardcoded old version of Pandas.

ERROR: Cannot install HuggingMouse, huggingmouse==0.2.0 and smolagents==1.7.0 because these package versions have conflicting dependencies.

The conflict is caused by:
huggingmouse 0.2.0 depends on pandas
allensdk 2.16.2 depends on pandas==1.5.3
smolagents 1.7.0 depends on pandas>=2.2.3

Any help would be appreciated!

Describe the solution you'd like
Would it be possible to update the Pandas version to newer?

Do you want to work on this issue?
I would be happy to delve into your code and help resolve the issue.

[RobertoDF]

+1 This is indeed a problem I also run into often

[yasithdev]

+1 indeed, ditto to all the version-constrained packages.

[mikejhuang]

I am the dev who set these dependencies.

Here's the PR relating to the setting of the dependencies
#2716

When testing pandas==2.0.3: it broke one or more of the tests with the following error, VisualBehaviorProjectCache.get_ophys_session_table( index_column="ophys_experiment_id").index is type 'object', expected int64

I don't believe there are officially any devs continuing to support this package but you are welcome to remove the dependency if the breaking change to that test does not affect your workflow or if you're able to fix the code that triggered the failure in that test.

[mariakesa]

Thank you mikejhuang! I am going to work on this in the near future.