Add support for numeric_only in DataFrame._reduce

[martinfalisse]

Add support for numeric_only in DataFrame._reduce, this way can use df.mean(numeric_only=True), etc. Resolves #2067. Also partially addresses #9009.

[GPUtester]

Can one of the admins verify this patch?

[GPUtester]

Can one of the admins verify this patch?

[quasiben]

ok to test

[codecov]

Codecov Report

Merging #10629 (b434124) into branch-22.06 (bc43e6a) will increase coverage by 0.05%.
The diff coverage is 91.40%.

❗ Current head b434124 differs from pull request most recent head 9ca5d12. Consider uploading reports for the commit 9ca5d12 to get more accurate results

@@               Coverage Diff                @@
##           branch-22.06   #10629      +/-   ##
================================================
+ Coverage         86.31%   86.37%   +0.05%     
================================================
  Files               140      142       +2     
  Lines             22280    22356      +76     
================================================
+ Hits              19232    19309      +77     
+ Misses             3048     3047       -1     

Impacted Files	Coverage Δ
python/cudf/cudf/core/algorithms.py	90.47% <ø> (ø)
python/cudf/cudf/core/groupby/groupby.py	91.72% <ø> (+0.22%)	⬆️
python/cudf/cudf/core/multiindex.py	92.14% <ø> (ø)
python/cudf/cudf/core/series.py	95.28% <ø> (ø)
python/cudf/cudf/utils/cudautils.py	59.83% <ø> (ø)
python/cudf/cudf/utils/ioutils.py	79.60% <ø> (ø)
python/cudf/cudf/utils/utils.py	90.28% <ø> (ø)
python/cudf/cudf/core/frame.py	93.67% <56.52%> (-1.09%)	⬇️
python/dask_cudf/dask_cudf/tests/utils.py	90.90% <90.90%> (ø)
python/cudf/cudf/core/column/lists.py	92.79% <100.00%> (+2.61%)	⬆️
... and 13 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update bc43e6a...9ca5d12. Read the comment docs.

[vyasr]

Thanks for the patch! I have a few small requests, then this should be good to go.

[vyasr]

Thanks, this LGTM!

[vyasr]

@gpucibot merge