evalml icon indicating copy to clipboard operation
evalml copied to clipboard

Error while running make_pipeline_from_data_check_output while applying data checks suggestion

Open sainiudit opened this issue 2 years ago • 1 comments

actions_pipeline = make_pipeline_from_data_check_output(problem_type, messages) data_df, y = actions_pipeline.fit(data_df, target)

################################################# Error Message: Screenshot 2023-01-25 at 8 18 04 PM

File /anaconda/envs/azureml_py310_sdkv2/lib/python3.10/site-packages/woodwork/logical_types.py:475, in IntegerNullable.transform(self, series, null_invalid_values) 473 if null_invalid_values: 474 series = _coerce_integer(series) --> 475 return super().transform(series)

File /anaconda/envs/azureml_py310_sdkv2/lib/python3.10/site-packages/woodwork/logical_types.py:76, in LogicalType.transform(self, series, null_invalid_values) 74 series = series.astype(new_dtype) 75 except (TypeError, ValueError): ---> 76 raise TypeConversionError(series, new_dtype, type(self)) 77 return series

TypeConversionError: Error converting datatype for LoanAmount from type float64 to type Int64. Please confirm the underlying data is consistent with logical type IntegerNullable.

sainiudit avatar Jan 25 '23 14:01 sainiudit

version of library:

import evalml, woodwork print("evalml : ",evalml.version) print("woodwork : ",woodwork.version)

evalml : 0.64.0 woodwork : 0.21.2

sainiudit avatar Jan 25 '23 14:01 sainiudit