For what I could see, there is still some functionality deprecated before 0.20.0, but which hasn't been removed, as it was missing from #6581.

This includes among others: * Categorical.labels, deprecated in 0.18.0 * The whole pandas.stats (rolling_mean...), deprecated in 0.18.0 * pandas.core.datetools, deprecated in 0.19.0 * flavor parameter of `SQLiteDatabase

I wrote a script that lists the version in which every FutureWarning was released (I think it makes sense adding it to pandas).

This is the output for versions older than 0.20.0:

v0.18.0 3ef82867 600) pandas/core/categorical.py v0.18.0 14cf67fb 132) pandas/core/resample.py v0.18.0 14cf67fb 685) pandas/core/resample.py v0.18.0 14cf67fb 700) pandas/core/resample.py v0.18.0 67730ddf 685) pandas/core/strings.py v0.18.0 3156395c 200) pandas/stats/moments.py v0.18.0 3156395c 240) pandas/stats/moments.py v0.18.1 e4df1ac7 2391) pandas/util/testing.py v0.19.0 3f3839b6 17) pandas/core/datetools.py v0.19.0 3f3839b6 67) pandas/util/_depr_module.py v0.19.0 8acfad34 47) pandas/io/sql.py v0.19.0 8acfad34 56) pandas/io/sql.py v0.19.0 95541952 101) pandas/util/_depr_module.py v0.19.0 b1bfd2fd 939) pandas/io/parsers.py

Can you please confirm if am I right, and I'll submit a PR?

Comment From: jreback

  • Categorical.labels, deprecated in 0.18.0
  • The whole pandas.stats (rolling_mean...), deprecated in 0.18.0

are already covered in #6581 (1st is available to do, 2nd is pending)

Comment From: jreback

added the remaining 2. pls feel free to do a PR for the ones that are available in #6581

Comment From: jreback

I wrote a script that lists the version in which every FutureWarning was released (I think it makes sense adding it to pandas).

sure would take this!

Comment From: datapythonista

Not sure if we also want to add to the list the expand parameter of Series.str.extract, which was set in 0.18.0 to change its default value from False to True with a deprecation warning.

Then, if I understood you correctly, everything I reported except pandas.stats can be removed, right?