from sklearn.preprocessing import StandardScaler scaler = StandardScaler(with_mean=False) scaler.fit_transform(X) What is the effect of setting with_mean=False?

Question

Data Science — Hard

from sklearn.preprocessing import StandardScaler scaler = StandardScaler(with_mean=False) scaler.fit_transform(X) What is the effect of setting with_mean=False?

Accepted Answer

StandardScaler centers data by subtracting the mean. For sparse matrices, subtracting the mean would destroy the sparsity by creating non-zero values for all zero entries, which would lead to memory overflow. Setting `with_mean=False` skips the centering step, preserving the sparsity of the input matrix while still performing unit variance scaling.