fb_prophet_parallel.py

"""Prophet by Facebook for TimeSeries with an example of parameter mutation."""
import importlib
import datatable as dt
import numpy as np
from h2oaicore.models import CustomTimeSeriesModel
from h2oaicore.systemutils import make_experiment_logger, loggerinfo, loggerwarning, loggerdebug
from h2oaicore.systemutils import (
    small_job_pool, save_obj, load_obj, user_dir, remove, config, max_threads
)
from h2oaicore.systemutils_more import arch_type
import os
import pandas as pd
import shutil
import random
import uuid
from sklearn.preprocessing import LabelEncoder, MinMaxScaler


class suppress_stdout_stderr(object):
    def __init__(self):
        self.null_fds = [os.open(os.devnull, os.O_RDWR) for x in range(2)]
        self.save_fds = [os.dup(1), os.dup(2)]

    def __enter__(self):
        os.dup2(self.null_fds[0], 1)
        os.dup2(self.null_fds[1], 2)

    def __exit__(self, *_):
        os.dup2(self.save_fds[0], 1)
        os.dup2(self.save_fds[1], 2)
        for fd in self.null_fds + self.save_fds:
            os.close(fd)


# Parallel implementation requires methods being called from different processes
# Global methods support this feature
# We use global methods as a wrapper for member methods of the transformer
def MyParallelProphetTransformer_fit_async(*args, **kwargs):
    return FBProphetParallelModel._fit_async(*args, **kwargs)


def MyParallelProphetTransformer_transform_async(*args, **kwargs):
    return FBProphetParallelModel._transform_async(*args, **kwargs)


class FBProphetParallelModel(CustomTimeSeriesModel):
    _regression = True
    _binary = False
    _multiclass = False
    _display_name = "FB_Prophet_Parallel"
    _description = "Facebook Prophet TimeSeries forecasting with multi process support"
    _parallel_task = True
    _testing_can_skip_failure = False  # ensure tested as if shouldn't fail

    @staticmethod
    def is_enabled():
        return not (arch_type == "ppc64le")

    @staticmethod
    def can_use(accuracy, interpretability, **kwargs):
        return False  # by default too slow unless only enabled

    @staticmethod
    def do_acceptance_test():
        return False

    froms3 = True
    if froms3:
        _root_path = "https://s3.amazonaws.com/artifacts.h2o.ai/deps/dai/recipes"
        _suffix = "-cp311-cp311-linux_x86_64.whl"
        _modules_needed_by_name = [
            '%s/setuptools_git-1.2%s' % (_root_path, _suffix),
            '%s/LunarCalendar-0.0.9%s' % (_root_path, _suffix),
            '%s/ephem-4.1.5%s' % (_root_path, _suffix),
            '%s/cmdstanpy-1.2.2%s' % (_root_path, _suffix),
            '%s/pystan-3.9.1%s' % (_root_path, _suffix),
            '%s/httpstan-4.12.0%s' % (_root_path, _suffix),
            '%s/prophet-1.1.5%s' % (_root_path, _suffix),
        ]
    else:
        _modules_needed_by_name = ['holidays==0.47', 'convertdate', 'lunarcalendar', 'pystan==3.9.1',
                                   'prophet==1.1.5']

    def set_default_params(self,
                           accuracy=None, time_tolerance=None, interpretability=None,
                           **kwargs):

        """
        Parameters available for the model :
          - growth : available market growth strategy in Prophet are linear and logistic
            logistic growth require a cap that saturates the predictions output
            See : https://facebook.github.io/prophet/docs/saturating_forecasts.html#forecasting-growth

          - country_holidays : allows Prophet to use built in Holidays
            See mutate_params to check the available countries in the model
            https://facebook.github.io/prophet/docs/seasonality,_holiday_effects,_and_regressors.html#built-in-country-holidays

          We can change the way seasonality affects the predictions
          - seasonality_mode : 'additive' (default) or 'multiplicative'

          We can override Fourier Order for seasonality calculation
          https://facebook.github.io/prophet/docs/seasonality,_holiday_effects,_and_regressors.html#fourier-order-for-seasonalities
          - weekly_seasonality : default is 'auto'
            Can be False or any number that gives the Fourier Order for the seasonality calculation
          - yearly_seasonality : default is 'auto'
            Can be False or any number that gives the Fourier Order for the seasonality calculation

          By default only weekly and yearly seasonality are calculated
          However one can ask Prophet to calculate other/specific seasonality
          https://facebook.github.io/prophet/docs/seasonality,_holiday_effects,_and_regressors.html#specifying-custom-seasonalities
          - monthly_seasonality : Either False (no monthly seasonality) or a number which will be the Fourier Order
            for monthly seasonality.

          - quarterly_seasonality : Either False (no quarterly seasonality) or a number which will be the Fourier Order
            for quarterly seasonality.
        """
        self.params = dict(
            growth=kwargs.get("growth", "linear"),
            seasonality_mode=kwargs.get("seasonality_mode", "additive"),
            country_holidays=kwargs.get("country_holidays", None),
            weekly_seasonality=kwargs.get("weekly_seasonality", 'auto'),
            monthly_seasonality=kwargs.get("monthly_seasonality", False),
            quarterly_seasonality=kwargs.get("quarterly_seasonality", False),
            yearly_seasonality=kwargs.get("yearly_seasonality", 'auto'),
        )

    def mutate_params(self,
                      accuracy, time_tolerance, interpretability,
                      **kwargs):

        logger = None
        if self.context and self.context.experiment_id:
            logger = make_experiment_logger(experiment_id=self.context.experiment_id, tmp_dir=self.context.tmp_dir,
                                            experiment_tmp_dir=self.context.experiment_tmp_dir)

        # Default version is do no mutation
        # Otherwise, change self.params for this model
        holiday_choice = [None, "US", "UK", "DE", "FRA"]
        if accuracy >= 8:
            weekly_choice = [False, 'auto', 5, 7, 10, 15]
            yearly_choice = [False, 'auto', 5, 10, 15, 20, 30]
            monthly_choice = [False, 3, 5, 7, 10]
            quarterly_choice = [False, 3, 5, 7, 10]
        elif accuracy >= 5:
            weekly_choice = [False, 'auto', 10, 20]
            yearly_choice = [False, 'auto', 10, 20]
            monthly_choice = [False, 5]
            quarterly_choice = [False, 5]
        else:
            # No alternative seasonality, and no seasonality override for weekly and yearly
            weekly_choice = [False, 'auto']
            yearly_choice = [False, 'auto']
            monthly_choice = [False]
            quarterly_choice = [False]

        self.params["country_holidays"] = np.random.choice(holiday_choice)
        self.params["seasonality_mode"] = np.random.choice(["additive", "multiplicative"])
        self.params["weekly_seasonality"] = np.random.choice(weekly_choice)
        self.params["monthly_seasonality"] = np.random.choice(monthly_choice)
        self.params["quarterly_seasonality"] = np.random.choice(quarterly_choice)
        self.params["yearly_seasonality"] = np.random.choice(yearly_choice)
        self.params["growth"] = np.random.choice(["linear", "logistic"])

    @staticmethod
    def _fit_async(X_path, grp_hash, tmp_folder):
        """
        Fits a FB Prophet model for a particular time group
        :param X_path: Path to the data used to fit the FB Prophet model
        :param grp_hash: Time group identifier
        :return: time group identifier and path to the pickled model
        """
        np.random.seed(1234)
        random.seed(1234)
        X = load_obj(X_path)
        # Commented for performance, uncomment for debug
        # print("prophet - fitting on data of shape: %s for group: %s" % (str(X.shape), grp_hash))
        if X.shape[0] < 20:
            # print("prophet - small data work-around for group: %s" % grp_hash)
            return grp_hash, None
        # Import FB Prophet package
        mod = importlib.import_module('prophet')
        Prophet = getattr(mod, "Prophet")
        nrows = X[['ds', 'y']].shape[0]
        n_changepoints = max(1, int(nrows * (2 / 3)))
        if n_changepoints < 25:
            model = Prophet(n_changepoints=n_changepoints)
        else:
            model = Prophet()

        with suppress_stdout_stderr():
            model.fit(X[['ds', 'y']])
        model_path = os.path.join(tmp_folder, "fbprophet_model" + str(uuid.uuid4()))
        save_obj(model, model_path)
        remove(X_path)  # remove to indicate success
        return grp_hash, model_path

    @staticmethod
    def _get_n_jobs(logger, **kwargs):
        if 'n_jobs_prophet' in config.recipe_dict:
            return min(config.recipe_dict['n_jobs_prophet'], max_threads())
        try:
            if config.fixed_num_folds <= 0:
                n_jobs = max(1, int(int(max_threads() / min(config.num_folds, kwargs['max_workers']))))
            else:
                n_jobs = max(1, int(
                    int(max_threads() / min(config.fixed_num_folds, config.num_folds, kwargs['max_workers']))))
        except KeyError:
            loggerinfo(logger, "Prophet No Max Worker in kwargs. Set n_jobs to 1")
            n_jobs = 1

        return n_jobs if n_jobs > 1 else 1

    def _clean_tmp_folder(self, logger, tmp_folder):
        try:
            shutil.rmtree(tmp_folder)
            loggerinfo(logger, "Prophet cleaned up temporary file folder.")
        except:
            loggerwarning(logger, "Prophet could not delete the temporary file folder.")

    def _create_tmp_folder(self, logger):
        # Create a temp folder to store files used during multi processing experiment
        # This temp folder will be removed at the end of the process
        # Set the default value without context available (required to pass acceptance test
        tmp_folder = os.path.join(user_dir(), "%s_prophet_model_folder" % uuid.uuid4())
        # Make a real tmp folder when experiment is available
        if self.context and self.context.experiment_id:
            tmp_folder = os.path.join(self.context.experiment_tmp_dir, "%s_prophet_model_folder" % uuid.uuid4())

        # Now let's try to create that folder
        try:
            os.mkdir(tmp_folder)
        except PermissionError:
            # This not occur so log a warning
            loggerwarning(logger, "Prophet was denied temp folder creation rights")
            tmp_folder = os.path.join(user_dir(), "%s_prophet_model_folder" % uuid.uuid4())
            os.mkdir(tmp_folder)
        except FileExistsError:
            # We should never be here since temp dir name is expected to be unique
            loggerwarning(logger, "Prophet temp folder already exists")
            tmp_folder = os.path.join(self.context.experiment_tmp_dir, "%s_prophet_model_folder" % uuid.uuid4())
            os.mkdir(tmp_folder)
        except:
            # Revert to temporary file path
            tmp_folder = os.path.join(user_dir(), "%s_prophet_model_folder" % uuid.uuid4())
            os.mkdir(tmp_folder)

        loggerinfo(logger, "Prophet temp folder {}".format(tmp_folder))
        return tmp_folder

    @staticmethod
    def _fit_async(X_path, grp_hash, tmp_folder, params, cap):
        """
        Fits a FB Prophet model for a particular time group
        :param X_path: Path to the data used to fit the FB Prophet model
        :param grp_hash: Time group identifier
        :return: time group identifier and path to the pickled model
        """
        np.random.seed(1234)
        random.seed(1234)
        X = load_obj(X_path)
        # Commented for performance, uncomment for debug
        # print("prophet - fitting on data of shape: %s for group: %s" % (str(X.shape), grp_hash))
        if X.shape[0] < 20:
            return grp_hash, None

        # Import FB Prophet package
        mod = importlib.import_module('prophet')
        Prophet = getattr(mod, "Prophet")

        # Fit current model and prior
        nrows = X[['ds', 'y']].shape[0]
        n_changepoints = max(1, int(nrows * (2 / 3)))
        if n_changepoints < 25:
            model = Prophet(growth=params["growth"], n_changepoints=n_changepoints)
        else:
            model = Prophet(growth=params["growth"])
        # Add params
        if params["country_holidays"] is not None:
            model.add_country_holidays(country_name=params["country_holidays"])
        if params["monthly_seasonality"]:
            model.add_seasonality(name='monthly', period=30.5, fourier_order=params["monthly_seasonality"])
        if params["quarterly_seasonality"]:
            model.add_seasonality(name='quarterly', period=92, fourier_order=params["quarterly_seasonality"])

        with suppress_stdout_stderr():
            if params["growth"] == "logistic":
                X["cap"] = cap
                model.fit(X[['ds', 'y', 'cap']])
            else:
                model.fit(X[['ds', 'y']])

        model_path = os.path.join(tmp_folder, "fbprophet_model" + str(uuid.uuid4()))
        save_obj(model, model_path)
        remove(X_path)  # remove to indicate success
        return grp_hash, model_path

    def get_hash(self, key):
        # Create dict key to store the min max scaler
        if isinstance(key, tuple):
            key = list(key)
        elif isinstance(key, list):
            pass
        else:
            # Not tuple, not list
            key = [key]
        grp_hash = '_'.join(map(str, key))
        return grp_hash

    def fit(self, X, y, sample_weight=None, eval_set=None, sample_weight_eval_set=None, **kwargs):

        # Get TGC and time column
        self.tgc = self.params_base.get('tgc', None)
        self.time_column = self.params_base.get('time_column', None)
        self.nan_value = np.mean(y)
        self.cap = np.max(y) * 1.5  # TODO Don't like this we should compute a cap from average yearly growth
        self.prior = np.mean(y)

        if self.time_column is None:
            self.time_column = self.tgc[0]

        # Get the logger if it exists
        logger = None
        if self.context and self.context.experiment_id:
            logger = make_experiment_logger(
                experiment_id=self.context.experiment_id,
                tmp_dir=self.context.tmp_dir,
                experiment_tmp_dir=self.context.experiment_tmp_dir
            )

        loggerinfo(logger, "Start Fitting Prophet Model with params : {}".format(self.params))

        try:
            # Add value of prophet_top_n in recipe_dict variable inside of config.toml file
            # eg1: recipe_dict="{'prophet_top_n': 200}"
            # eg2: recipe_dict="{'prophet_top_n':10}"
            self.top_n = config.recipe_dict['prophet_top_n']
        except KeyError:
            self.top_n = 50

        loggerinfo(logger, f"Prophet will use {self.top_n} groups as well as average target data.")

        # Get temporary folders for multi process communication
        tmp_folder = self._create_tmp_folder(logger)

        n_jobs = self._get_n_jobs(logger, **kwargs)

        # Reduce X to TGC
        tgc_wo_time = list(np.setdiff1d(self.tgc, self.time_column))
        X = X[:, self.tgc].to_pandas()

        # Fill NaNs or None
        X = X.replace([None, np.nan], 0)

        # Add target, Label encoder is only used for Classif. which we don't support...
        if self.labels is not None:
            y = LabelEncoder().fit(self.labels).transform(y)
        X['y'] = np.array(y)

        self.nan_value = X['y'].mean()

        # Change date feature name to match Prophet requirements
        X.rename(columns={self.time_column: "ds"}, inplace=True)

        # Create a general scale now that will be used for unknown groups at prediction time
        # Can we do smarter than that ?
        general_scaler = MinMaxScaler().fit(X[['y', 'ds']].groupby('ds').median().values)

        # Go through groups and standard scale them
        if len(tgc_wo_time) > 0:
            X_groups = X.groupby(tgc_wo_time)
        else:
            X_groups = [([None], X)]

        scalers = {}
        scaled_ys = []

        print('Number of groups : ', len(X_groups))
        for g in tgc_wo_time:
            print(f'Number of groups in {g} groups : {X[g].unique().shape}')

        for key, X_grp in X_groups:
            # Create dict key to store the min max scaler
            grp_hash = self.get_hash(key)
            # Scale target for current group
            scalers[grp_hash] = MinMaxScaler()
            y_skl = scalers[grp_hash].fit_transform(X_grp[['y']].values)
            # Put back in a DataFrame to keep track of original index
            y_skl_df = pd.DataFrame(y_skl, columns=['y'])

            y_skl_df.index = X_grp.index
            scaled_ys.append(y_skl_df)

        # Set target back in original frame but keep original
        X['y_orig'] = X['y']
        X['y'] = pd.concat(tuple(scaled_ys), axis=0)

        # Now Average groups
        X_avg = X[['ds', 'y']].groupby('ds').mean().reset_index()

        # Send that to Prophet
        mod = importlib.import_module('prophet')
        Prophet = getattr(mod, "Prophet")
        nrows = X[['ds', 'y']].shape[0]
        n_changepoints = max(1, int(nrows * (2 / 3)))
        if n_changepoints < 25:
            model = Prophet(yearly_seasonality=True, weekly_seasonality=True, daily_seasonality=True,
                            n_changepoints=n_changepoints)
        else:
            model = Prophet(yearly_seasonality=True, weekly_seasonality=True, daily_seasonality=True)

        if self.params["country_holidays"] is not None:
            model.add_country_holidays(country_name=self.params["country_holidays"])
        if self.params["monthly_seasonality"]:
            model.add_seasonality(name='monthly', period=30.5, fourier_order=self.params["monthly_seasonality"])
        if self.params["quarterly_seasonality"]:
            model.add_seasonality(name='quarterly', period=92, fourier_order=self.params["quarterly_seasonality"])

        with suppress_stdout_stderr():
            model.fit(X[['ds', 'y']])

        top_groups = None
        if len(tgc_wo_time) > 0:
            if self.top_n > 0:
                top_n_grp = X.groupby(tgc_wo_time).size().sort_values().reset_index()[tgc_wo_time].iloc[
                            -self.top_n:].values
                top_groups = [
                    '_'.join(map(str, key))
                    for key in top_n_grp
                ]

        grp_models = {}
        priors = {}
        if top_groups:
            # Prepare for multi processing
            num_tasks = len(top_groups)

            def processor(out, res):
                out[res[0]] = res[1]

            pool_to_use = small_job_pool
            loggerinfo(logger, f"Prophet will use {n_jobs} workers for fitting.")

            pool = pool_to_use(
                logger=None, processor=processor,
                num_tasks=num_tasks, max_workers=n_jobs
            )
            #
            # Fit 1 FB Prophet model per time group columns
            nb_groups = len(X_groups)

            # Put y back to its unscaled value for top groups
            X['y'] = X['y_orig']

            for _i_g, (key, X) in enumerate(X_groups):
                # Just log where we are in the fitting process
                if (_i_g + 1) % max(1, nb_groups // 20) == 0:
                    loggerinfo(logger, "FB Prophet : %d%% of groups fitted" % (100 * (_i_g + 1) // nb_groups))

                X_path = os.path.join(tmp_folder, "fbprophet_X" + str(uuid.uuid4()))
                X = X.reset_index(drop=True)
                save_obj(X, X_path)

                grp_hash = self.get_hash(key)

                if grp_hash not in top_groups:
                    continue

                priors[grp_hash] = X['y'].mean()

                args = (X_path, grp_hash, tmp_folder, self.params, self.cap)
                kwargs = {}
                pool.submit_tryget(None, MyParallelProphetTransformer_fit_async,
                                   args=args, kwargs=kwargs, out=grp_models)
            pool.finish()

            for k, v in grp_models.items():
                grp_models[k] = load_obj(v) if v is not None else None
                remove(v)

        self._clean_tmp_folder(logger, tmp_folder)

        self.set_model_properties(
            model={
                'avg': model,
                'group': grp_models,
                'priors': priors,
                'topgroups': top_groups,
                'skl': scalers,
                'gen_scaler': general_scaler
            },
            features=self.tgc,  # Prophet uses time and timegroups
            importances=np.ones(len(self.tgc)),
            iterations=-1  # Does not have iterations
        )

        return None

    @staticmethod
    def _transform_async(model_path, X_path, nan_value, tmp_folder):
        """
        Predicts target for a particular time group
        :param model_path: path to the stored model
        :param X_path: Path to the data used to fit the FB Prophet model
        :param nan_value: Value of target prior, used when no fitted model has been found
        :return: self
        """
        model = load_obj(model_path)
        XX_path = os.path.join(tmp_folder, "fbprophet_XXt" + str(uuid.uuid4()))
        X = load_obj(X_path)
        # Facebook Prophet returns the predictions ordered by time
        # So we should keep track of the time order for each group so that
        # predictions are ordered the same as the imput frame
        # Keep track of the order
        order = np.argsort(pd.to_datetime(X["ds"]))
        if model is not None:
            # Run prophet
            yhat = model.predict(X)['yhat'].values
            XX = pd.DataFrame(yhat, columns=['yhat'])
        else:
            XX = pd.DataFrame(np.full((X.shape[0], 1), nan_value), columns=['yhat'])  # invalid models
        XX.index = X.index[order]
        assert XX.shape[1] == 1
        save_obj(XX, XX_path)
        remove(model_path)  # indicates success, no longer need
        remove(X_path)  # indicates success, no longer need
        return XX_path

    def predict(self, X: dt.Frame, **kwargs):
        """
        Uses fitted models (1 per time group) to predict the target
        :param X: Datatable Frame containing the features
        :return: FB Prophet predictions
        """
        # Get the logger if it exists
        logger = None
        if self.context and self.context.experiment_id:
            logger = make_experiment_logger(
                experiment_id=self.context.experiment_id,
                tmp_dir=self.context.tmp_dir,
                experiment_tmp_dir=self.context.experiment_tmp_dir
            )

        if self.tgc is None or not all([x in X.names for x in self.tgc]):
            loggerdebug(logger, "Return 0 predictions")
            return np.ones(X.shape[0]) * self.nan_value

        models, _, _, _ = self.get_model_properties()

        model = models['avg']
        grp_models = models['group']
        priors = models['priors']
        top_groups = models['topgroups']
        scalers = models['skl']
        general_scaler = models['gen_scaler']

        tmp_folder = self._create_tmp_folder(logger)

        n_jobs = self._get_n_jobs(logger, **kwargs)

        # Reduce X to TGC
        tgc_wo_time = list(np.setdiff1d(self.tgc, self.time_column))
        X = X[:, self.tgc].to_pandas()

        # Fill NaNs or None
        X = X.replace([None, np.nan], 0)

        # Change date feature name to match Prophet requirements
        X.rename(columns={self.time_column: "ds"}, inplace=True)

        if self.params["growth"] == "logistic":
            X["cap"] = self.cap

        # Predict y using unique dates
        X_time = X[['ds']].groupby('ds').first().reset_index()
        with suppress_stdout_stderr():
            y_avg = model.predict(X_time)[['ds', 'yhat']]

        # Prophet transforms the date column to datetime so we need to transfrom that to merge back
        X_time.sort_values('ds', inplace=True)
        X_time['yhat'] = y_avg['yhat']
        X_time.sort_index(inplace=True)

        # Merge back into original frame on 'ds'
        # pd.merge wipes the index ... so keep it to provide it again
        indices = X.index
        X = pd.merge(
            left=X,
            right=X_time[['ds', 'yhat']],
            on='ds',
            how='left'
        )
        X.index = indices

        # Go through groups and recover the scaled target for knowed groups
        if len(tgc_wo_time) > 0:
            X_groups = X.groupby(tgc_wo_time)
        else:
            X_groups = [([None], X)]

        inverted_ys = []
        for key, X_grp in X_groups:
            grp_hash = self.get_hash(key)

            # Scale target for current group
            if grp_hash in scalers.keys():
                inverted_y = scalers[grp_hash].inverse_transform(X_grp[['yhat']])
            else:
                inverted_y = general_scaler.inverse_transform(X_grp[['yhat']])

            # Put back in a DataFrame to keep track of original index
            inverted_df = pd.DataFrame(inverted_y, columns=['yhat'])
            inverted_df.index = X_grp.index
            inverted_ys.append(inverted_df)

        XX_general = pd.concat(tuple(inverted_ys), axis=0).sort_index()

        if top_groups:
            # Go though the groups and predict only top
            XX_paths = []
            model_paths = []

            def processor(out, res):
                out.append(res)

            num_tasks = len(top_groups)
            pool_to_use = small_job_pool
            pool = pool_to_use(logger=None, processor=processor, num_tasks=num_tasks, max_workers=n_jobs)

            nb_groups = len(X_groups)
            for _i_g, (key, X_grp) in enumerate(X_groups):

                # Just log where we are in the fitting process
                if (_i_g + 1) % max(1, nb_groups // 20) == 0:
                    loggerinfo(logger, "FB Prophet : %d%% of groups predicted" % (100 * (_i_g + 1) // nb_groups))

                # Create dict key to store the min max scaler
                grp_hash = self.get_hash(key)
                X_path = os.path.join(tmp_folder, "fbprophet_Xt" + str(uuid.uuid4()))

                if grp_hash not in top_groups:
                    XX = pd.DataFrame(np.full((X_grp.shape[0], 1), np.nan), columns=['yhat'])  # unseen groups
                    XX.index = X_grp.index
                    save_obj(XX, X_path)
                    XX_paths.append(X_path)
                    continue

                if grp_models[grp_hash] is None:
                    XX = pd.DataFrame(np.full((X_grp.shape[0], 1), np.nan), columns=['yhat'])  # unseen groups
                    XX.index = X_grp.index
                    save_obj(XX, X_path)
                    XX_paths.append(X_path)
                    continue

                model = grp_models[grp_hash]
                model_path = os.path.join(tmp_folder, "fbprophet_modelt" + str(uuid.uuid4()))
                save_obj(model, model_path)
                save_obj(X_grp, X_path)
                model_paths.append(model_path)

                args = (model_path, X_path, priors[grp_hash], tmp_folder)
                kwargs = {}
                pool.submit_tryget(None, MyParallelProphetTransformer_transform_async, args=args, kwargs=kwargs,
                                   out=XX_paths)

            pool.finish()
            XX_top_groups = pd.concat((load_obj(XX_path) for XX_path in XX_paths), axis=0).sort_index()
            for p in XX_paths + model_paths:
                remove(p)

        self._clean_tmp_folder(logger, tmp_folder)

        features_df = pd.DataFrame()
        features_df['GrpAvg'] = XX_general['yhat']

        if top_groups:
            features_df[f'_Top{self.top_n}Grp'] = XX_top_groups['yhat']
            features_df.loc[
                features_df[f'_Top{self.top_n}Grp'].notnull(), 'GrpAvg'
            ] = features_df.loc[
                features_df[f'_Top{self.top_n}Grp'].notnull(), f'_Top{self.top_n}Grp'
            ]

        # Models have to return a numpy array
        return features_df['GrpAvg'].values