def fit_predict()

in orbit/diagnostics/backtest.py [0:0]

• 48 lines of code
• 4 McCabe index (conditional complexity)

    def fit_predict(self):
        """Fit and predict on each data split and set predicted_df

        Since this part of the backtesting is generally the most expensive, BackTester
        breaks up fit/predict and scoring into two separate calls

        Returns
        -------
        None

        """
        splitter = self._splitter
        model = self.model
        response_col = model.response_col
        date_col = model.date_col
        output_res = list()
        for train_df, test_df, scheme, key in splitter.split():
            model_copy = deepcopy(model)
            model_copy.fit(train_df)
            train_predictions = model_copy.predict(train_df)
            test_predictions = model_copy.predict(test_df)
            all_pred_cols = [x for x in train_predictions.columns if x != date_col]

            # set attributes
            self._fitted_models.append(model_copy)
            self._splitter_scheme.append(scheme)
            self._test_actual = np.concatenate(
                (self._test_actual, test_df[response_col].to_numpy())
            )
            self._test_prediction = np.concatenate(
                (
                    self._test_prediction,
                    test_predictions[BacktestFitKeys.PREDICTED.value].to_numpy(),
                )
            )
            self._train_actual = np.concatenate(
                (self._train_actual, train_df[response_col].to_numpy())
            )
            self._train_prediction = np.concatenate(
                (
                    self._train_prediction,
                    train_predictions[BacktestFitKeys.PREDICTED.value].to_numpy(),
                )
            )

            # set df attribute
            # join train
            train_dates = train_df[date_col].rename(BacktestFitKeys.DATE.value)
            train_response = train_df[response_col].rename(BacktestFitKeys.ACTUAL.value)
            train_values = pd.concat(
                (train_dates, train_response, train_predictions[all_pred_cols]), axis=1
            )
            train_values[BacktestFitKeys.TRAIN_FLAG.value] = True
            # join test
            test_dates = test_df[date_col].rename(BacktestFitKeys.DATE.value)
            test_response = test_df[response_col].rename(BacktestFitKeys.ACTUAL.value)
            test_values = pd.concat(
                (test_dates, test_response, test_predictions[all_pred_cols]), axis=1
            )
            test_values[BacktestFitKeys.TRAIN_FLAG.value] = False
            # union train/test
            both_values = pd.concat((train_values, test_values), axis=0)
            both_values[BacktestFitKeys.SPLIT_KEY.value] = key
            output_res.append(both_values)
        # union each splits
        self._predicted_df = pd.concat(output_res, axis=0).reset_index(drop=True)