DEPR: Enforce 2.0 deprecations

[datapythonista]

xref #30228

List of deprecations pending enforcement before releasing pandas 2.0.

Topic	Deprecation PR	Enforcement PR	Assignee
datetimelike.astype(...)	#38544 #45574	#49715	@jbrockmendel
Int64Index, UInt64Index & Float64Index	#42717 #43028	#49560	@topper-123
DataFrame.to_latex using Styler	#47970		@attack68

Those are from unchecked items in #30228 and #50196 (comment). We should probably run a grep to see what else is pending before we start merging PRs that introduce new deprecations (see #50578).

[jbrockmendel]

Is this redundant with #30228?

[mroeschke]

Yeah it appears these are essentially the "last" depercations that need to be enforced in #30228

[rhshadrach]

#35967 can be removed since it was never merged, yea?

[datapythonista]

Is this redundant with #30228?

Yes, I thought it'd make it easier to track things. But on a second thought, I think it's probably better to simply have the list of blockers in #46776, whether they're a deprecation or not. So, closing here in favor of the RLS 2.0 issue. Sorry for the noise.

[topper-123]

What is the status of #35967 wrt. changing the default for the observed in groupby? The parameter default was never deprecated, so that means it will not change for 2.0?

This issue is something that relatively often bites people I think, so would be good to change, either as a deprecation or a 2.0 change.

[topper-123]

To add: #35967 is the issue listed in #30228, #43999 is a the same issue, but it's open, so may be the one intended for discussion about this.

[rhshadrach]

This issue is something that relatively often bites people I think, so would be good to change, either as a deprecation or a 2.0 change.

I'm planning to push for deprecating in 2.x, but would be -1 on changing in 2.0 itself.