I am also seeing this warning on the console when running rsmtool:

Setting an item of incompatible dtype is deprecated and will raise in a future error of pandas. Value '-' has dtype incompatible with int64, please explicitly cast to a compatible dtype first.

@tamarl08 we need another issue right?

Codecov Report

Patch coverage is 100.00% of modified lines.

📢 Thoughts on this report? Let us know!.

May be worth running all of the tools to check any other warnings in the reports and on the console, while we are at it?

@desilinguist

• I ran test with STRICT locally on this branch and it passes, does it check the console tools?
• I'll open an issue for followup on seaboarn
• I see these warning too. I was not 100% sure if they are new. should we resolve them before release too?

1. No, that just checks warnings in reports.
2. Great!
3. Definitely new.

@desilinguist I fixed the warning by setting these cells to None - this is better than -1 because then the cell in the table is empty which is nicer.
rsmexplain also has a new warning which is fixed in shap and will be included in their next version. I opened an issue for that too.

Perfect! I think the n_init warning has been there for a while. Good to know that they are fixing it.

Hmm, if I test with the lr-exclude-flags-and-zeros test, the table cell has a NaN in it, it's not empty. Also shouldn't the numbers all be integers instead of some being floats?

@tamarl08 speaking of warnings, I am seeing a couple of new ones when I run the lr-exclude-flags-and-zeros:

DataFrame.applymap has been deprecated. Use DataFrame.map instead.
DataFrame.applymap has been deprecated. Use DataFrame.map instead.

Perfect! I think the n_init warning has been there for a while. Good to know that they are fixing it.

Hmm, if I test with the lr-exclude-flags-and-zeros test, the table cell has a NaN in it, it's not empty. Also shouldn't the numbers all be integers instead of some being floats?

Which report is this? I don't see this table.

Just found another warning in summarize report:

Series.__getitem__ treating keys as positions is deprecated. In a future version, integer keys will always be treated as labels (consistent with DataFrame behavior). To access a value by position, use `ser.iloc[pos]`

Perfect! I think the n_init warning has been there for a while. Good to know that they are fixing it.

Hmm, if I test with the lr-exclude-flags-and-zeros test, the table cell has a NaN in it, it's not empty. Also shouldn't the numbers all be integers instead of some being floats?

Which report is this? I don't see this table.

It's rsmtool using the test I mentioned. Run that manually.