def train_production_models(best_models):
"""
Prepares all the production models
:param best_models:
:return: list of 4 models lists with trained models
"""
hogwarts_df = load_processed_data()
slitherin_model = train_single_production_model(hogwarts_df, best_models[0], 'is_slitherin')
griffindor_model = train_single_production_model(hogwarts_df, best_models[1], 'is_griffindor')
ravenclaw_model = train_single_production_model(hogwarts_df, best_models[2], 'is_ravenclaw')
hufflpuff_model = train_single_production_model(hogwarts_df, best_models[3], 'is_hufflpuff')
return slitherin_model, griffindor_model, ravenclaw_model, hufflpuff_model
def train_all_models():
"""
Trains models for each faculty.
:return: list of 4 models lists with trained models
"""
hogwarts_df = load_processed_data()
slitherin_models = get_faculty_models(hogwarts_df, 'is_slitherin')
griffindor_models = get_faculty_models(hogwarts_df, 'is_griffindor')
ravenclaw_models = get_faculty_models(hogwarts_df, 'is_ravenclaw')
hufflpuff_models = get_faculty_models(hogwarts_df, 'is_hufflpuff')
return slitherin_models, griffindor_models, ravenclaw_models, hufflpuff_models
def train_production_models(best_models):
"""
Prepares all the production models
:param best_models:
:return: list of 4 models lists with trained models
"""
fict_df = load_processed_data()
fem_model = train_single_production_model(fict_df, best_models[0],
'is_fem')
fict_model = train_single_production_model(fict_df, best_models[1],
'is_fict')
fim_model = train_single_production_model(fict_df, best_models[2],
'is_fim')
gef_model = train_single_production_model(fict_df, best_models[3],
'is_gef')
return fem_model, fict_model, fim_model, gef_model
def train_all_models():
"""
Trains models for each faculty.
:return: list of 4 models lists with trained models
"""
fict_df = load_processed_data()
fem_model = train_single_production_model(fict_df, best_models[0],
'is_fem')
fict_model = train_single_production_model(fict_df, best_models[1],
'is_fict')
fim_model = train_single_production_model(fict_df, best_models[2],
'is_fim')
gef_model = train_single_production_model(fict_df, best_models[3],
'is_gef')
return fem_model, fict_model, fim_model, gef_model
from notebooks.model_training import train_classifiers
from data_loaders import load_processed_data
import warnings
warnings.filterwarnings('ignore')
# Загружаем данные
fict_df = load_processed_data()
# Оставляем только нужные колонки
data_full = fict_df.drop(['name', 'surname', 'is_fict', 'is_fim', 'is_gef'],
axis=1).copy()
X_data = data_full.drop('is_fem', axis=1)
y = data_full.is_fem
# Проводим исследование моделей
print("123")
fem_models = train_classifiers(X_data, y)
score_testing_dataset(fem_models[5])