This will reduce the chances of bugs due to accidentally mutating a data frame: https://pandas.pydata.org/docs/dev/user_guide/copy_on_write.html