Implemented GroupBy.median()

[itholic]

This PR proposes GroupBy.median().

Note: the result can be slightly different from pandas since we use an approximated median based upon approximate percentile computation because computing median across a large dataset is extremely expensive.

>>> kdf = ks.DataFrame({'a': [1., 1., 1., 1., 2., 2., 2., 3., 3., 3.],
...                     'b': [2., 3., 1., 4., 6., 9., 8., 10., 7., 5.],
...                     'c': [3., 5., 2., 5., 1., 2., 6., 4., 3., 6.]},
...                    columns=['a', 'b', 'c'],
...                    index=[7, 2, 4, 1, 3, 4, 9, 10, 5, 6])
>>> kdf
      a     b    c
7   1.0   2.0  3.0
2   1.0   3.0  5.0
4   1.0   1.0  2.0
1   1.0   4.0  5.0
3   2.0   6.0  1.0
4   2.0   9.0  2.0
9   2.0   8.0  6.0
10  3.0  10.0  4.0
5   3.0   7.0  3.0
6   3.0   5.0  6.0

>>> kdf.groupby('a').median().sort_index()  # doctest: +NORMALIZE_WHITESPACE
       b    c
a
1.0  2.0  3.0
2.0  8.0  2.0
3.0  7.0  4.0

>>> kdf.groupby('a')['b'].median().sort_index()
a
1.0    2.0
2.0    8.0
3.0    7.0
Name: b, dtype: float64

ref #1929

[itholic]

FYI: we're doing same thing in median for DataFrame and Series.

koalas/databricks/koalas/generic.py

Lines 1898 to 1904 in 9e8d99b

	def median(self, axis=None, numeric_only=True, accuracy=10000) -> Union[Scalar, "Series"]:
	"""
	Return the median of the values for the requested axis.
	.. note:: Unlike pandas', the median in Koalas is an approximated median based upon
	approximate percentile computation because computing median across a large dataset
	is extremely expensive.

[codecov-io]

Codecov Report

Merging #1957 (e121472) into master (a68717d) will increase coverage by 0.00%.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##           master    #1957   +/-   ##
=======================================
  Coverage   94.59%   94.60%           
=======================================
  Files          49       49           
  Lines       10882    10890    +8     
=======================================
+ Hits        10294    10302    +8     
  Misses        588      588           

Impacted Files	Coverage Δ
databricks/koalas/missing/groupby.py	100.00% <ø> (ø)
databricks/koalas/groupby.py	91.58% <100.00%> (+0.07%)	⬆️
databricks/koalas/series.py	96.90% <0.00%> (+0.01%)	⬆️

---

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 a68717d...e121472. Read the comment docs.

[xinrong-meng]

Otherwise, LGTM!

[ueshin]

Could you fix the conflicts as well?
Otherwise, LGTM.

[ueshin]

Are numeric_only and accuracy unknown types?

[ueshin]

We should add more type hints for function arguments in the separate PR.

[itholic]

Sure, I'll add in the separated PR.

Thanks for the review!

[itholic]

Oh, anyway numeric_only only supports for True since this is only for pandas compatibility.

Should we add type hints as bool though ?

[ueshin]

Yep, it must be bool, and accuracy must be int?
There are other similar places in the file or the other files, we can do it in a batch.

[itholic]

Sounds good. I will take a look and add type hints throughout the whole file next week !

[ueshin]

Thanks! I'd merge this now.
Let's add more type hints for function arguments later.