orbit/forecaster/full_bayes.py [151:166]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
            point_predicted_dict = self._model.predict(
                posterior_estimates=point_posteriors,
                df=df,
                training_meta=training_meta,
                prediction_meta=prediction_meta,
                include_error=False,
                **kwargs,
            )
            for k, v in point_predicted_dict.items():
                point_predicted_dict[k] = np.squeeze(v, 0)

            # to derive confidence interval; the condition should be sufficient since we add [50] by default
            if self._n_bootstrap_draws > 0 and len(self._prediction_percentiles) > 1:
                # perform bootstrap; we don't have posterior samples. hence, we just repeat the draw here.
                posterior_samples = {}
                for k, v in point_posteriors.items():
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -



orbit/forecaster/map.py [55:71]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
        point_predicted_dict = self._model.predict(
            posterior_estimates=point_posteriors,
            df=df,
            training_meta=training_meta,
            prediction_meta=prediction_meta,
            # false for point estimate
            include_error=False,
            **kwargs,
        )
        for k, v in point_predicted_dict.items():
            point_predicted_dict[k] = np.squeeze(v, 0)

        # to derive confidence interval; the condition should be sufficient since we add [50] by default
        if self._n_bootstrap_draws > 0 and len(self._prediction_percentiles) > 1:
            # perform bootstrap; we don't have posterior samples. hence, we just repeat the draw here.
            posterior_samples = {}
            for k, v in point_posteriors.items():
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -