-
-
Notifications
You must be signed in to change notification settings - Fork 18.8k
Closed
Labels
AlgosNon-arithmetic algos: value_counts, factorize, sorting, isin, clip, shift, diffNon-arithmetic algos: value_counts, factorize, sorting, isin, clip, shift, diffMissing-datanp.nan, pd.NaT, pd.NA, dropna, isnull, interpolatenp.nan, pd.NaT, pd.NA, dropna, isnull, interpolate
Milestone
Description
xref #19938. There's I'm abusing the fact that the Int64HashTable class treats iNaT as a missing value. It'd be cleaner to just pass the Categorical's codes with a flag saying that -1
is the missing value marker.
IIUC, the missing value condition is baked into the class definition.
{{py:
# name, dtype, null_condition, float_group
dtypes = [('Float64', 'float64', 'val != val', True),
('UInt64', 'uint64', 'False', False),
('Int64', 'int64', 'val == iNaT', False)]
I'm not familiar enough with Cython to know how that could be made a parameter. Possibly an optional scalar value. If that value is not None check if val == na_sentinel
. Else, use the condition from the class definition?
Metadata
Metadata
Assignees
Labels
AlgosNon-arithmetic algos: value_counts, factorize, sorting, isin, clip, shift, diffNon-arithmetic algos: value_counts, factorize, sorting, isin, clip, shift, diffMissing-datanp.nan, pd.NaT, pd.NA, dropna, isnull, interpolatenp.nan, pd.NaT, pd.NA, dropna, isnull, interpolate