There might be cases where you're willing to accept the error, would like to go on and just suppress the warning message (example: moving average of a column that starts with zeros...).

In case you just want to suppress these warnings, just write

np.seterr(invalid='ignore')

after the numpy import statement.