def assert_frame_sort_equal(a, b, **kwargs):
"""Tests that DataFrames are equal, even if rows are in different order"""
df_a = ungroup(a)
df_b = ungroup(b)
sorted_a = robust_multiple_sort(df_a, list(df_a.columns)).reset_index(drop = True)
sorted_b = robust_multiple_sort(df_b, list(df_b.columns)).reset_index(drop = True)
assert_frame_equal(sorted_a, sorted_b, **kwargs)
def assert_frame_sort_equal(a, b):
"""Tests that DataFrames are equal, even if rows are in different order"""
df_a = ungroup(a)
df_b = ungroup(b)
sorted_a = df_a.sort_values(by = df_a.columns.tolist()).reset_index(drop = True)
sorted_b = df_b.sort_values(by = df_b.columns.tolist()).reset_index(drop = True)
assert_frame_equal(sorted_a, sorted_b)
def assert_equal_query(tbl, lazy_query, target, **kwargs):
out = collect(lazy_query(tbl))
if isinstance(tbl, pd.DataFrame):
df_a = ungroup(out).reset_index(drop=True)
df_b = ungroup(target).reset_index(drop=True)
assert_frame_equal(df_a, df_b, **kwargs)
else:
assert_frame_sort_equal(out, target, **kwargs)
def test_dply_grouped_mutate_of_agg_order():
# see issue #139
df = pd.DataFrame({
'g': ['b', 'a', 'b'],
'x':[0, 1, 2]
})
gdf = df.groupby('g')
out = mutate(gdf, g_min = lambda d: d.x.min())
assert_frame_equal(ungroup(out), df.assign(g_min = [0, 1, 0]))