Update Polars dtype test to generate more examples

[deepyaman]

Otherwise, it's always testing all-null dataframes, as far as I can tell (see https://discord.com/channels/897120336003334214/1160076741394563102/1268358149400690719).

Also dropped the include_nulls keyword, which I don't think does anything (see pola-rs/polars#17969).

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 93.31%. Comparing base (812b2a8) to head (a454729).
Report is 154 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1770      +/-   ##
==========================================
- Coverage   94.28%   93.31%   -0.97%     
==========================================
  Files          91      120      +29     
  Lines        7013     9129    +2116     
==========================================
+ Hits         6612     8519    +1907     
- Misses        401      610     +209     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[alexander-beedie]

FYI: I wouldn't necessarily drop include_nulls completely. Instead, I'd consider replacing it with the updated param allow_null=True:

allow_null : bool or Mapping[str, bool]
    Allow nulls as possible values and allow the `Null` data type by default.
    Accepts either a boolean or a mapping of column names to booleans.

[deepyaman]

FYI: I wouldn't necessarily drop include_nulls completely. Instead, I'd consider replacing it with the updated param allow_null=True:

allow_null : bool or Mapping[str, bool]
    Allow nulls as possible values and allow the `Null` data type by default.
    Accepts either a boolean or a mapping of column names to booleans.

Sorry, I missed this before; I'm OK either way, as the argument defaults to True anyway.