"model": XGBRegressor(),
"params": {
"gamma": np.random.uniform(low=0.01, high=0.05, size=10),
"max_depth": [4, 5, 6],
"min_child_weight": [4, 5, 6],
"reg_alpha": [1e-5, 1e-2, 0.1, 1, 10, 100]
}
}
}
if __name__ == "__main__":
df = load_sol_challenge()
# Data Preprocessing
preprocessor = PreProcessor()
df = preprocessor.str_to_float(df, cols_with_str)
df = preprocessor.remove_nans(df)
# EDA
# Data Distribution
data_distribution = DataDistribution(cols_to_analyse,
PATH_RESULTS_EDA_DIST,
ignore_outliers=False)
data_distribution.run(df)
# Feature Correlation
feature_correlation = FeatureCorrelation(cols_to_analyse,
PATH_RESULTS_EDA_CORR,
figsize=(9, 9))
feature_correlation.run(df)
# Get independent and dependent variables
X = np.asarray(df[X_names_num])
