While doing some experiments with the library, my inputs were not accepted and I was getting 0 features found errors. I did some investigation and discovered a bug: My dtypes were all uint8 but it is not listed in the selected dtypes.
Previously only these dtypes are allowed as inputs: ["int64", "float64", "int32", "float32"]. However, there are more than that. Official pandas documentation states that to select all number dtypes, np.number can be used. This PR implements this.
While doing some experiments with the library, my inputs were not accepted and I was getting 0 features found errors. I did some investigation and discovered a bug: My dtypes were all uint8 but it is not listed in the selected dtypes.
Previously only these dtypes are allowed as inputs:
["int64", "float64", "int32", "float32"]
. However, there are more than that. Official pandas documentation states that to select all number dtypes,np.number
can be used. This PR implements this.Thanks.