def create_datasources()

in k-fold-cross-validation/fold.py [0:0]

• 35 lines of code
• 3 McCabe index (conditional complexity)

    def create_datasources(self):
        """
        Creates datasource for model training and evaluation on Amazon ML.
        """
        # create training datasource for this fold
        self.train_ds_id = "ds-" + base64.b32encode(os.urandom(10)).decode(
            "ascii")
        self.train_ds_rearrange = self.build_rearrangement_str(
            is_complement=True)
        self.train_ds_name = self.build_datasource_name(
            self.data_spec.name, self.train_ds_rearrange)

        self._ml.create_data_source_from_s3(
            data_source_id=self.train_ds_id,
            data_source_name=self.train_ds_name,
            data_spec={
                "DataLocationS3": self.data_spec.data_s3_url,
                "DataSchema": self.data_spec.schema,
                "DataRearrangement": self.train_ds_rearrange
            },
            compute_statistics=True
        )
        logger.info("Created Training Datasource " + self.train_ds_id)

        # create evaluation datasource for this fold
        self.eval_ds_id = "ds-" + base64.b32encode(os.urandom(10)).decode(
            "ascii")
        self.eval_ds_rearrange = self.build_rearrangement_str(
            is_complement=False)
        self.eval_ds_name = self.build_datasource_name(
            self.data_spec.name, self.eval_ds_rearrange)
        self._ml.create_data_source_from_s3(
            data_source_id=self.eval_ds_id,
            data_source_name=self.eval_ds_name,
            data_spec={
                "DataLocationS3": self.data_spec.data_s3_url,
                "DataSchema": self.data_spec.schema,
                "DataRearrangement": self.eval_ds_rearrange
            },
            compute_statistics=True
        )
        logger.info("Created Evaluation Datasource " + self.eval_ds_id)