0

I have read the documentation: https://pandas.pydata.org/docs/reference/api/pandas.DataFrame.values.html

My df has a mix of strings, objects and int dtypes. I have not found a way to control for using .values() on my df without having to reconvert my previous int dtype elements back from the downcast dtype.

Is there an easier way to do this?

Clay Campbell
  • 168
  • 13
  • What is your goal? –  Feb 02 '22 at 16:17
  • `values` points to the underlying numpy array, which necessarily has a unique dtype. What do you want to do here? Have you tried `.to_records()`? – mozway Feb 02 '22 at 16:21

0 Answers0