Replace np.float with np.float64 in one_hot_encoder

[sarahyurick]

Any background context you want to provide?

NumPy 1.24.0 removed the np.float alias, so we have to replace np.float with np.float64 in one_hot_encoder.

Another option is to replace np.float with Python's float, although the NumPy version may be useful for consistency with NumPy arrays (source).

[beckernick]

cc @weixuanfu , is this something that TPOT would be open to updating to support recent NumPy version?

[gkirgizov]

Vote for this PR, a required change. Please, merge this

[joseortiz3]

Please, someone with write access merge this. Every time I install tpot I have to change this line (I change it to float and that seems to work as well).

[sarahyurick]

Hi @jay-m-dev or @perib I was wondering if you have the permissions to review and merge this?