itholic commented on code in PR #37801:
URL: https://github.com/apache/spark/pull/37801#discussion_r963156365
##########
python/pyspark/pandas/groupby.py:
##########
@@ -895,6 +895,95 @@ def sem(col: Column) -> Column:
bool_to_numeric=True,
)
+ # TODO: 1, 'n' accepts list and slice; 2, implement 'dropna' parameter
Review Comment:
Maybe do we want to create a ticket as a sub-tasks of SPARK-40327 ?
##########
python/pyspark/pandas/groupby.py:
##########
@@ -895,6 +895,89 @@ def sem(col: Column) -> Column:
bool_to_numeric=True,
)
+ # TODO: 1, 'n' accepts list and slice; 2, implement 'dropna' parameter
+ def nth(self, n: int) -> FrameLike:
+ """
+ Take the nth row from each group.
+
+ .. versionadded:: 3.4.0
+
+ Parameters
+ ----------
+ n : int
+ A single nth value for the row
+
+ Examples
+ --------
+
+ >>> df = ps.DataFrame({'A': [1, 1, 2, 1, 2],
+ ... 'B': [np.nan, 2, 3, 4, 5]}, columns=['A', 'B'])
+ >>> g = df.groupby('A')
+ >>> g.nth(0)
+ B
+ A
+ 1 NaN
+ 2 3.0
+ >>> g.nth(1)
+ B
+ A
+ 1 2.0
+ 2 5.0
+ >>> g.nth(-1)
+ B
+ A
+ 1 4.0
+ 2 5.0
+
+ See Also
+ --------
+ pyspark.pandas.Series.groupby
+ pyspark.pandas.DataFrame.groupby
+ """
+ groupkey_names = [SPARK_INDEX_NAME_FORMAT(i) for i in
range(len(self._groupkeys))]
+ internal, agg_columns, sdf = self._prepare_reduce(
+ groupkey_names=groupkey_names,
+ accepted_spark_types=None,
+ bool_to_numeric=False,
+ )
+ psdf: DataFrame = DataFrame(internal)
+
+ if len(psdf._internal.column_labels) > 0:
+ window1 =
Window.partitionBy(*groupkey_names).orderBy(NATURAL_ORDER_COLUMN_NAME)
+ tmp_row_number_col = "__tmp_row_number_col__"
+ if n >= 0:
+ sdf = (
+ psdf._internal.spark_frame.withColumn(
+ tmp_row_number_col, F.row_number().over(window1)
+ )
+ .where(F.col(tmp_row_number_col) == n + 1)
+ .drop(tmp_row_number_col)
+ )
+ else:
+ window2 = Window.partitionBy(*groupkey_names).rowsBetween(
+ Window.unboundedPreceding, Window.unboundedFollowing
+ )
+ tmp_group_size_col = "__tmp_group_size_col__"
+ sdf = (
+ psdf._internal.spark_frame.withColumn(
+ tmp_group_size_col, F.count(F.lit(0)).over(window2)
+ )
+ .withColumn(tmp_row_number_col,
F.row_number().over(window1))
+ .where(F.col(tmp_row_number_col) ==
F.col(tmp_group_size_col) + 1 + n)
+ .drop(tmp_group_size_col, tmp_row_number_col)
+ )
+ else:
+ sdf = sdf.select(*groupkey_names).distinct()
Review Comment:
If there is a bug in pandas, maybe we should add a test by manually creating
the expected result rather than just skipping the test ?
e.g.
https://github.com/apache/spark/blob/6d2ce128058b439094cd1dd54253372af6977e79/python/pyspark/pandas/tests/test_series.py#L1654-L1658
##########
python/pyspark/pandas/groupby.py:
##########
@@ -895,6 +895,95 @@ def sem(col: Column) -> Column:
bool_to_numeric=True,
)
+ # TODO: 1, 'n' accepts list and slice; 2, implement 'dropna' parameter
+ def nth(self, n: int) -> FrameLike:
+ """
+ Take the nth row from each group.
+
+ .. versionadded:: 3.4.0
+
+ Parameters
+ ----------
+ n : int
+ A single nth value for the row
+
+ Returns
+ -------
+ Series or DataFrame
+
+ Examples
+ --------
+ >>> df = ps.DataFrame({'A': [1, 1, 2, 1, 2],
+ ... 'B': [np.nan, 2, 3, 4, 5]}, columns=['A', 'B'])
+ >>> g = df.groupby('A')
+ >>> g.nth(0)
+ B
+ A
+ 1 NaN
+ 2 3.0
+ >>> g.nth(1)
+ B
+ A
+ 1 2.0
+ 2 5.0
+ >>> g.nth(-1)
+ B
+ A
+ 1 4.0
+ 2 5.0
+
+ See Also
+ --------
+ pyspark.pandas.Series.groupby
+ pyspark.pandas.DataFrame.groupby
+ """
+ if not isinstance(n, int):
+ raise TypeError("Unsupported type %s" % type(n).__name__)
Review Comment:
nit: Maybe rather raises `NotImplementedError` since we should support the
other types in the future ?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
