More details on pandas dtypes

[timmens]

Describe the inconsistency / error

In the screencast on (pandas) dtypes, I've noticed that

1. The Float32Dtype is not mentioned (on the "Overview of numeric dtypes" slide)
2. The Float64Dtype does not have a range (on the "Overview of numeric dtypes" slide)
3. The BooleanDtype is not mentioned
4. The DatetimeTZDtype is not mentioned

I'd argue that

• The Float32 dtype should be added to the list for consistency
• The ranges of the single and double precision floats could be added as info
• The boolean and datetime dtype could be shown on a "Miscellaneous" slide

Location

• https://effective-programming-practices.vercel.app/pandas/datatypes/objectives_materials.html
• https://github.com/OpenSourceEconomics/epp_topics/blob/main/src/epp_topics/pandas/datatypes/screencast/slides.md

[janosg]

I agree that bool and Datetime would be good to add.

For the float dtypes I would recommend they always use 64. That's why I did not mention anything else. For floats it's less about range and more about the significant digits you can get.