ThomasLecocq
commented
4 years ago do these two behave differently? df.mean() vs df.apply(np.mean) to avoid using df.values.mean()
Open ianozsvald opened 4 years ago
ThomasLecocq
commented
4 years ago do these two behave differently? df.mean() vs df.apply(np.mean) to avoid using df.values.mean()
A quick test with %timeit suggests that an Int64 is 4x slower than an int64...but this isn't conclusive. Since an Int64 doesn't have a .values I can't use my usual loop so this requires a bit more work.