def main():
# read in package config
config = configparser.ConfigParser()
config.read("config/config.yml")
# read in input data
all_data = read_input_data.get_data(config["read_input_data_config"])
# preprocess input data
data_preprocessor = preprocessing.DataPreprocessor(all_data)
data_preprocessor.preprocess_data()
# run hyperparameter search
hyp_param_search = total_grid_search.HypParamSearch(
data_preprocessor.X_data,
data_preprocessor.y_data,
col_na_proportion=data_preprocessor.col_na_proportion)
grid_search_cv_preprocess = hyp_param_search.grid_hyp_search(
config["grid_search_config"])
bayes_search_cv_preprocess = hyp_param_search.bayes_opt_hyp_search(
config["bayes_opt_search_config"])
# print hyperparameter search results
total_grid_search.get_grid_search_results(grid_search_cv_preprocess)
total_grid_search.get_grid_search_results(bayes_search_cv_preprocess)
def setup_test(self):
self.config = configparser.ConfigParser()
self.config.read("config/config.yml")
self.config = self.config["read_input_data_config"]
self.data = read_input_data.get_data(self.config)
self.col_names = read_input_data.get_col_names(self.config["col_names_file_path"])
self.col_names.append("data_origin")
self.col_types = read_input_data.read_json_to_dict(self.config["col_types_file_path"])
def setup_test(self):
# read in package config
config = configparser.ConfigParser()
config.read("config/config.yml")
# read in input data
all_data = read_input_data.get_data(config["read_input_data_config"])
# preprocess input data
self.data_preprocessor = DataPreprocessor(all_data)
self.data_preprocessor.preprocess_data()
def setup_test(self):
self.config = configparser.ConfigParser()
self.config.read("config/config.yml")
self.read_input_config = self.config["read_input_data_config"]
self.data = read_input_data.get_data(self.read_input_config)
self.data_preprocessor = DataPreprocessor(self.data)
self.data_preprocessor.preprocess_data()
self.hyp_param_search = HypParamSearch(
self.data_preprocessor.X_data,
self.data_preprocessor.y_data,
col_na_proportion=self.data_preprocessor.col_na_proportion)
def setup_test(self):
self.config = configparser.ConfigParser()
self.config.read("config/config.yml")
self.config = self.config["read_input_data_config"]
self.data = read_input_data.get_data(self.config)
data_preprocessor = DataPreprocessor(self.data)
data_preprocessor.preprocess_data()
self.feature_selector = FeatureSelector(
X_data=data_preprocessor.X_data,
y_data=data_preprocessor.y_data,
col_na_proportion=data_preprocessor.col_na_proportion,
missing_value_filter=0.5,
feature_importance_rank_filter=20)
def setup_test(self):
self.config = configparser.ConfigParser()
self.config.read("config/config.yml")
self.config = self.config["read_input_data_config"]
self.data = read_input_data.get_data(self.config)
self.data_preprocessor = DataPreprocessor(self.data)
self.data_preprocessor.preprocess_data()
self.total_estimator = TotalEstimator(missing_value_filter=0.5,
feature_importance_rank_filter=5,
n_estimators=200,
max_features=8,
max_leaf_nodes=25,
fit_params=None,
times_fit=1)
self.X = self.data_preprocessor.X_data
self.y = self.data_preprocessor.y_data
fit_params = {
"col_na_proportion": self.data_preprocessor.col_na_proportion
}
self.fitted_estimator = self.total_estimator.fit(
self.X, self.y, **fit_params)