[SPARK-40621][PS] Implement numeric_only and min_count in GroupBy.sum

[zhengruifeng]

What changes were proposed in this pull request?

Implement numeric_only and min_count in GroupBy.sum

Why are the changes needed?

for API coverage

Does this PR introduce any user-facing change?

new parameter

In [2]: df = ps.DataFrame({"A": [1, 2, 1, 2], "B": [True, False, False, True], "C": [3, 4, 3, 4], "D": ["a", "a", "b", "a"]})

In [3]: df.groupby("A").sum(numeric_only=False).sort_index()
/Users/ruifeng.zheng/Dev/spark/python/pyspark/pandas/utils.py:975: PandasAPIOnSparkAdviceWarning: GroupBy.sum() can only support numeric and bool columns even ifnumeric_only=False, skip unsupported columns: ['D']
  warnings.warn(message, PandasAPIOnSparkAdviceWarning)
                                                                                
   B  C
A      
1  1  6
2  1  8

In [4]: df._to_pandas().groupby("A").sum(numeric_only=False).sort_index()
Out[4]: 
   B  C   D
A          
1  1  6  ab
2  1  8  aa

In [5]: df.groupby("D").sum(min_count=3).sort_index()
Out[5]: 
     A    B     C
D                
a  5.0  2.0  11.0
b  NaN  NaN   NaN

In [6]: df._to_pandas().groupby("D").sum(min_count=3).sort_index()
Out[6]: 
     A    B     C
D                
a  5.0  2.0  11.0
b  NaN  NaN   NaN

How was this patch tested?

added UT

[zhengruifeng]

given a non-numeric column, for example, str type, the final result is sensitive to the order, so not easy to implement for now.
Right now, warn the users that such columns will be skiped:
PandasAPIOnSparkAdviceWarning: GroupBy.sum() can only support numeric and bool columns even ifnumeric_only=False, skip unsupported columns: ['D']

[HyukjinKwon]

Merged to master.

[zhengruifeng]

@HyukjinKwon thank you for reviewing!