python - aws boto3 attributes 100% missing values -
when create datasource
via python script below, @ least 1 of attributes has 100% missing values. when manually create datasource
via aws ml dashboard, , apply same attribute types, none of values missing. there problem how i'm creating datasource s3?
file_names = [file_name_train, file_name_testing] client = boto3.client('machinelearning') schema_file = open('../selections/aws_schema.txt', 'r') schema = schema_file.read() file_name in file_names: response = client.create_data_source_from_s3( datasourceid=file_name+date, datasourcename=file_name+date, dataspec={ 'datalocations3': 's3://'+bucket_name+'/'+file_name+file_extension, 'dataschema': schema, }, computestatistics=true )
Comments
Post a Comment