Variance of nullable Int64 ignores ddof argument · Issue #29128 · pandas-dev/pandas (original) (raw)

Variance calculation Series.var() is ignoring the ddof argument for nullable integer type Int64. Behavior should match that of float type.

pd.version '0.25.1' np.version '1.17.2' s1 = pd.Series(data=[1,2,3,4,5,6,np.nan,np.nan]) series.var(ddof=0) 3.5 s1 = pd.Series(data=[1,2,3,4,5,6,np.nan,np.nan]) s1.var(ddof=0) 2.9166666666666665 s1.var(ddof=1) 3.5 s2 = pd.Series(data=[1,2,3,4,5,6,np.nan,np.nan], dtype="Int64") s2.var(ddof=0) 3.5 s2.var(ddof=1) 3.5