7 ms·
Right, each pandas column is a numpy array (or optionally pyarrow in pandas 2.0), but numpy is going to be faster for things that needs to be vectorized across
by anywhichway 3y ago
Right, each pandas column is a numpy array (or optionally pyarrow in pandas 2.0), but numpy is going to be faster for things that needs to be vectorized across columns, such as matrix operations.
- jamann 3y agoThats when you do polars.DataFrame.to_numpy()