search

ieaves / tenzing

MIT License
0 stars 0 forks source link

Integer / Float Logic Flawed for int64 #25

Closed ieaves closed 5 years ago

ieaves commented 5 years ago

Int64 allows pandas to support np.NaN's as part of an integer dtype. The current implementation does not capture this dynamic and should be fixed.

ieaves commented 5 years ago

Also an issue with infs; integer series containging infs are cast to float64 by pandas and not currently tested for by tenzing_float.