Commit 07125ea

authored

DOC: add select_dtypes case to 3.0 string migration guide (#62403)

1 parent 831d29d commit 07125eaCopy full SHA for 07125ea

File tree

+12

-0

lines changed

+12

-0

lines changed

Lines changed: 12 additions & 0 deletions

Original file line number	Diff line number	Diff line change
`@@ -195,6 +195,18 @@ This is actually compatible with pandas 2.x as well, since in pandas < 3,`
`195`	`195`	`of also stringifying missing values in pandas 2.x. See the section`
`196`	`196`	:ref:`string_migration_guide-astype_str` for more details.
`197`	`197`
	`198`	+For selecting string columns with :meth:`~DataFrame.select_dtypes` in a pandas
	`199`	+2.x and 3.x compatible way, it is not possible to use ``"str"``. While this
	`200`	`+works for pandas 3.x, it raises an error in pandas 2.x.`
	`201`	+As an alternative, you can select both ``object`` (for pandas 2.x) and
	`202`	+``"string"`` (for pandas 3.x; which will also select the default ``str`` dtype
	`203`	`+and does not error on pandas 2.x):`
	`204`	`+`
	`205`	`+.. code-block:: python`
	`206`	`+`
	`207`	+ # can use ``include=["str"]`` for pandas >= 3
	`208`	`+ >>> df.select_dtypes(include=["object", "string"])`
	`209`	`+`
`198`	`210`
`199`	`211`	`The missing value sentinel is now always NaN`
`200`	`212`	`~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~`

Comments

(0)