SimpleImputer should support String type

Hi,

I’m getting this error:

RuntimeError: Operator SklearnSimpleImputer (type: SklearnSimpleImputer) got an input merged_columns with a wrong type <class 'onnxconverter_common.data_types.StringTensorType'>. Only [<class 'onnxconverter_common.data_types.FloatTensorType'>, <class 'onnxconverter_common.data_types.Int64TensorType'>, <class 'onnxconverter_common.data_types.DoubleTensorType'>] are allowed

But when using the option strategy="most_frequent" or strategy="constant", it supports strings, see https://scikit-learn.org/stable/modules/generated/sklearn.impute.SimpleImputer.html.

[gs-alt]

I get the same error as @victornoel

[victornoel]

Hey @xadupre, it has been some time, and there is still one thing that does not work: when I use None as the value for missing_values:

SimpleImputer(missing_values=None, strategy='most_frequent')

I get this error:

NotImplementedError: The converter is implemented when the missing values are string not <class 'NoneType'>.