From fbeb0a86f1d3efc96263a81981b0a059d93fa4f5 Mon Sep 17 00:00:00 2001 From: Christophe Guyeux Date: Tue, 18 Feb 2020 11:26:52 +0100 Subject: [PATCH] Integrating historical features --- config/learn.cfg | 4 ++ main.py | 5 +- predictops/engine.py | 9 ++++ predictops/learn/preprocessing.py | 80 +++++++++++++++++++------------ 4 files changed, 64 insertions(+), 34 deletions(-) diff --git a/config/learn.cfg b/config/learn.cfg index f6b2e43..1a9566e 100644 --- a/config/learn.cfg +++ b/config/learn.cfg @@ -19,5 +19,9 @@ fill_method = spline order = 3 +[HISTORY_KNOWLEDGE] +nb_lines = 5 + + [TARGET] config = (Path.cwd() / 'config') / 'targets' / 'sdis25.cfg' diff --git a/main.py b/main.py index d1d7f9c..cf8fe81 100644 --- a/main.py +++ b/main.py @@ -1,5 +1,4 @@ from predictops.engine import Engine -from predictops.learn.preprocessing import Preprocessing from logging import getLogger from logging.config import fileConfig @@ -17,10 +16,8 @@ if __name__ == '__main__': engine.add_features() engine.add_target() - process = Preprocessing(config_file = config, dict_features = engine.X) + engine.add_preprocessing() - print(process.dataframe.head(n=20)) - print(process.dataframe.tail(n=20)) '''target = toarea(stream_file = Path.cwd() / 'data' / 'targets' / 'sdis25' / 'interventions.csv') diff --git a/predictops/engine.py b/predictops/engine.py index dedd265..44ab9c4 100644 --- a/predictops/engine.py +++ b/predictops/engine.py @@ -7,6 +7,7 @@ from shutil import rmtree from .source.ephemeris import Ephemeris from .source.meteofrance import MeteoFrance +from .learn.preprocessing import Preprocessing from .target.target import Target fileConfig((Path.cwd() / 'config') / 'logging.cfg') @@ -71,6 +72,14 @@ class Engine: timestep = self._timestep) + def add_preprocessing(self): + process = Preprocessing(config_file = self._config, + dict_features = self.X, + dict_target = self.y) + print(process.dataframe.head(n=2)) + + + @property def X(self): return self._X diff --git a/predictops/learn/preprocessing.py b/predictops/learn/preprocessing.py index 4197b8f..187a5b7 100644 --- a/predictops/learn/preprocessing.py +++ b/predictops/learn/preprocessing.py @@ -25,12 +25,12 @@ class Preprocessing: - NaN values are then filled with last known values. ''' - def __init__(self, config_file = None, dict_features = None, features = None): + def __init__(self, config_file = None, + dict_features = None, dict_target = None): ''' Constructor that defines all needed attributes and collects features. ''' - self._config = ConfigParser() - self._config.read(config_file) + self._config = config_file self._start = datetime.strptime(self._config['DATETIME']['start'], '%m/%d/%Y %H:%M:%S') @@ -39,16 +39,15 @@ class Preprocessing: self._timestep = timedelta(hours = self._config['DATETIME'].getfloat('hourStep')) self._dict_features = dict_features + self._dict_target = dict_target + self._full_dict = None self._dataframe = None self._datetimes = [] - # If features are not provided to the constructor, then we collect - # any existing feature in the dictionary - if features != None: - self._features = features - else: - self._features = set(chain.from_iterable([tuple(u.keys()) + + self._features = set(chain.from_iterable([tuple(u.keys()) for u in [*dict_features.values()]])) + feature_files = Path.cwd() / 'config' / 'features' self._features = {feat : {'numerical': False} for feat in self._features} for feature_file in listdir(feature_files): @@ -66,6 +65,12 @@ class Preprocessing: if config.has_option(section, 'numerical'): self._features[section]['numerical'] = config[section].getboolean('numerical') + self._numerical_columns = [k for k in self._features if self._features[k]['type'] == 1 + or (self._features[k]['type'] == 3 and self._features[k]['numerical'])] + + self._categorical_columns = [k for k in self._features if self._features[k]['type'] == 2 + or (self._features[k]['type'] == 3 and not self._features[k]['numerical'])] + @property @@ -141,27 +146,23 @@ class Preprocessing: ''' logger.info("Filling NaN numerical values in the feature dataframe") # We interpolate (linearly or with splines) only numerical columns - numerical_columns = [k for k in self._features if self._features[k]['type'] == 1 - or (self._features[k]['type'] == 3 and self._features[k]['numerical'])] # The interpolation if self._config['PREPROCESSING']['fill_method'] == 'propagate': - self._dataframe[numerical_columns] =\ - self._dataframe[numerical_columns].fillna(method='ffill') + self._dataframe[self._numerical_columns] =\ + self._dataframe[self._numerical_columns].fillna(method='ffill') elif self._config['PREPROCESSING']['fill_method'] == 'linear': - self._dataframe[numerical_columns] =\ - self._dataframe[numerical_columns].interpolate() + self._dataframe[self._numerical_columns] =\ + self._dataframe[self._numerical_columns].interpolate() elif self._config['PREPROCESSING']['fill_method'] == 'spline': - self._dataframe[numerical_columns] =\ - self._dataframe[numerical_columns].interpolate(method='spline', + self._dataframe[self._numerical_columns] =\ + self._dataframe[self._numerical_columns].interpolate(method='spline', order=self._config['PREPROCESSING'].getint('order')) # For the categorical columns, NaN values are filled by duplicating # the last known value (forward fill method) logger.info("Filling NaN categorical values in the feature dataframe") - categorical_columns = [k for k in self._features if self._features[k]['type'] == 2 - or (self._features[k]['type'] == 3 and not self._features[k]['numerical'])] - self._dataframe[categorical_columns] =\ - self._dataframe[categorical_columns].fillna(method='ffill') + self._dataframe[self._categorical_columns] =\ + self._dataframe[self._categorical_columns].fillna(method='ffill') # Uncomment this line to fill NaN values at the beginning of the # dataframe. This may not be a good idea, especially for features @@ -175,15 +176,33 @@ class Preprocessing: if k not in self._datetimes]) + def _add_history(self): + ''' + Integrating previous nb of interventions as features + ''' + logger.info("Integrating previous nb of interventions as features") + nb_lines = self._config['HISTORY_KNOWLEDGE'].getint('nb_lines') + print(len(self._dataframe)) + print(self._dataframe.head(4)) + for k in range(1,nb_lines+1): + name = 'history_'+str(nb_lines-k+1) + self._dataframe[name] = [np.NaN]*k + list(self._dict_target.values())[:-k] + self._numerical_columns.append(name) + self._dataframe = self._dataframe[nb_lines:] + print(self._dataframe.head(4)) + print(len(self._dataframe)) + + + def _standardize(self): ''' Normalizing numerical features ''' logger.info("Standardizing numerical values in the feature dataframe") # We operate only on numerical columns - numerical_columns = [k for k in self._features if self._features[k]['type'] == 1 - or (self._features[k]['type'] == 3 and self._features[k]['numerical'])] - self._dataframe[numerical_columns] = preprocessing.scale(self._dataframe[numerical_columns]) + self._dataframe[self._numerical_columns] =\ + preprocessing.scale(self._dataframe[self._numerical_columns]) + def _one_hot_encoding(self): @@ -191,18 +210,17 @@ class Preprocessing: Apply a one hot encoding for category features ''' logger.info("One hot encoding for categorical feature") - categorical_columns = [k for k in self._features if self._features[k]['type'] == 2 - or (self._features[k]['type'] == 3 and not self._features[k]['numerical'])] - # On fait un codage disjonctif complet des variables qualitatives + # We store numerical columns df_out = pd.DataFrame() - for col in categorical_columns: + for col in self._numerical_columns: + df_out[col] = self._dataframe[col] + # The one hot encoding + for col in self._categorical_columns: pd1 = pd.get_dummies(self._dataframe[col],prefix=col) for col1 in pd1.columns: df_out[col1] = pd1[col1] self._dataframe = df_out - print(self._dataframe.head()) - @property @@ -216,6 +234,8 @@ class Preprocessing: orient='index') # Dealing with NaN values self._fill_nan() + # Adding previous (historical) nb_interventions as features + self._add_history() # Normalizing numerical values self._standardize() # Dealing with categorical features -- 2.39.5