PERF-#6224: Use 'Map' operator to retrieve categorical codes

[dchigarev]

What do these changes do?

After #6222 we guarantee that each partition has knowledge about the whole categorical table, meaning that we don't need to build the whole axis no more in order to retrieve valid categorical codes. So this PR basically replaces .fold(axis=0, ...) to the .map(...) in the .cat_codes implementation.

• first commit message and PR title follow format outlined here

  NOTE: If you edit the PR title to match this format, you need to add another commit (even if it's empty) or amend your last commit for the CI job that checks the PR title to pick up the new PR title.

• passes flake8 modin/ asv_bench/benchmarks scripts/doc_checker.py
• passes black --check modin/ asv_bench/benchmarks scripts/doc_checker.py
• signed commit with git commit -s
• Resolves Implement cat.codes using map rather than full-axis approach #6224
• tests are passing
• module layout described at docs/development/architecture.rst is up-to-date

[dchigarev]

since now, we shouldn't do this cast as disappearance of the categorical dtype will mean that we lost valid codes and should report an error

[anmyachev]

LGTM! @YarShev?

[YarShev]

We should probably update the map's signature in modin/core/dataframe/base/dataframe/dataframe.py.

[dchigarev]

I did that, however, it seems that there's already a mismatch with the base map as it additionally takes an axis parameter that is not defined in the pandas version of the dataframe

modin/modin/core/dataframe/base/dataframe/dataframe.py

Lines 90 to 96 in 930b8fb

	@abstractmethod
	def map(
	self,
	function: Callable,
	axis: Optional[Union[int, Axis]] = None,
	dtypes: Optional[str] = None,
	) -> "ModinDataframe":

[YarShev]

Let's align the signatures in a separate issue, please create one for this.

[dchigarev]

done, #6231