Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Fix bug 2406: filter dictionary aggregation keys to limit them to keys only present in current partition #2407

Merged
merged 3 commits into from
Nov 14, 2020
Merged

Fix bug 2406: filter dictionary aggregation keys to limit them to keys only present in current partition #2407

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
414cd25
FIX-#2406: Added test to detect this bug
gshimansky Nov 13, 2020
966d3a4
FIX-#2406: Added filter for keys absent in current partition
gshimansky Nov 13, 2020
7ce895e
FIX-#2406: Attemt to fix broken test on BaseOnPython backend
gshimansky Nov 14, 2020
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
14 changes: 9 additions & 5 deletions modin/backends/pandas/query_compiler.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2606,11 +2606,15 @@ def groupby_agg_builder(df):
def compute_groupby(df):
grouped_df = df.groupby(by=by, axis=axis, **groupby_kwargs)
try:
result = (
grouped_df.agg(agg_func)
if isinstance(agg_func, dict)
else agg_func(grouped_df, **agg_kwargs)
)
if isinstance(agg_func, dict):
# Filter our keys that don't exist in this partition. This happens when some columns
# from this original dataframe didn't end up in every partition.
partition_dict = {
k: v for k, v in agg_func.items() if k in df.columns
}
result = grouped_df.agg(partition_dict)
else:
result = agg_func(grouped_df, **agg_kwargs)
# This happens when the partition is filled with non-numeric data and a
# numeric operation is done. We need to build the index here to avoid
# issues with extracting the index.
Expand Down
40 changes: 27 additions & 13 deletions modin/pandas/test/test_groupby.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -22,6 +22,7 @@
check_df_columns_have_nans,
create_test_dfs,
eval_general,
test_data,
test_data_values,
modin_df_almost_equals_pandas,
)
Expand Down Expand Up @@ -1189,23 +1190,36 @@ def test_shift_freq(groupby_axis, shift_axis):
)


def test_agg_func_None_rename():
pandas_df = pandas.DataFrame(
@pytest.mark.parametrize(
"by_and_agg_dict",
[
{
"col1": np.random.randint(0, 100, size=1000),
"col2": np.random.randint(0, 100, size=1000),
"col3": np.random.randint(0, 100, size=1000),
"col4": np.random.randint(0, 100, size=1000),
"by": [
list(test_data["int_data"].keys())[0],
list(test_data["int_data"].keys())[1],
],
"agg_dict": {
"max": (list(test_data["int_data"].keys())[2], np.max),
"min": (list(test_data["int_data"].keys())[2], np.min),
},
},
index=["row{}".format(i) for i in range(1000)],
)
modin_df = from_pandas(pandas_df)
{
"by": ["col1"],
"agg_dict": {
"max": (list(test_data["int_data"].keys())[0], np.max),
"min": (list(test_data["int_data"].keys())[-1], np.min),
},
},
],
)
def test_agg_func_None_rename(by_and_agg_dict):
modin_df, pandas_df = create_test_dfs(test_data["int_data"])

modin_result = modin_df.groupby(["col1", "col2"]).agg(
max=("col3", np.max), min=("col3", np.min)
modin_result = modin_df.groupby(by_and_agg_dict["by"]).agg(
**by_and_agg_dict["agg_dict"]
)
pandas_result = pandas_df.groupby(["col1", "col2"]).agg(
max=("col3", np.max), min=("col3", np.min)
pandas_result = pandas_df.groupby(by_and_agg_dict["by"]).agg(
**by_and_agg_dict["agg_dict"]
)
df_equals(modin_result, pandas_result)

Expand Down