import logging
import pandas as pd
import numpy as np
import os

from sklearn.externals import joblib

from sklearn.preprocessing import Imputer
from sklearn.impute import SimpleImputer


logger = logging.getLogger(__name__)


def streams(pop_5, playlists):
    """Apply streams model based on popularity after 5 days and playlist that the track exists on, 
    including the playlist position. 

    NOTE: there is a hardcoded path included here!
    
    Parameters
    ----------
    pop_5: int
        Popularity after 5 days
    playlists: dict( playlistId -> position on playlist )
        Playlists and their position

    Returns
    -------
    Predicted total streams after 5 days INCLUDING the first five days

    """
    
    # TODO: Fix this hardcoded path???
    # Version hardcodes NA play lists with 1000
    model_spec_file = os.path.join(os.path.dirname(__file__), 'models/tpot_20_20_xgboos_fill_na_1000.pkl')
    # model_spec_file =  'models/tpot_20_20_xgboos_fill_na_1000.pkl'

    model_spec = joblib.load(model_spec_file)
    pipe_line = model_spec['model']
    feature_columns = model_spec['feature_cols']
    imputer = SimpleImputer(strategy='constant', fill_value=1000)

    unknown_playlist_ids = [x for x in playlists.keys() if x not in feature_columns]

    if len(unknown_playlist_ids) > 0:
        logger.warning("Unknown playlists when modelling")

    df = pd.DataFrame(columns=feature_columns)

    playlists['pop_5'] = pop_5 # Hmm this should sit somewhere else

    values = pd.DataFrame({k: [v] for (k, v) in playlists.items() if k in feature_columns})

    in_df = pd.concat([df, values], sort=False)
    imputer.fit(in_df)
    in_df = imputer.transform(in_df)

    np.exp(pipe_line.predict(in_df))[0]
    return np.exp(pipe_line.predict(in_df))[0]