Estadísticas y predicciones de Davis Cup World Group 2 Main

Rafael Nadal (España)

Aunque no participe activamente este año, Nadal sigue siendo una inspiración para muchos jugadores españoles. Su legado sigue presente y motiva a nuevos talentos a alcanzar grandes logros dentro del circuito profesional. <|repo_name|>zzangpp/kaggle-competitions<|file_sep|>/house-prices-advanced-regression-techniques/README.md # House Prices Advanced Regression Techniques The goal of this competition is to predict the sales price of houses based on various features. ## Approach ### EDA #### Correlation Analysis First I did some correlation analysis on the data to see which features are most correlated with the target variable (`SalePrice`). Here's the result:  As expected `GrLivArea` (above grade living area) is highly correlated with `SalePrice`. Other strong correlations are with `GarageArea` and `TotalBsmtSF`. It's interesting to note that `OverallQual` (the overall material and finish quality) is also highly correlated with `SalePrice`. #### Visualizing Correlations To better visualize these correlations I plotted scatter plots for each feature against the target variable:  I also plotted histograms of each feature against the target variable:  These plots show that there are some outliers in the data that may need to be addressed. #### Missing Data Next I checked for missing data in the dataset:  There are quite a few columns with missing data. I decided to fill in missing values for numerical columns with the median value and for categorical columns with the mode value. ### Feature Engineering #### Creating New Features I created some new features based on existing ones: * `TotalSF` = `TotalBsmtSF` + `1stFlrSF` + `2ndFlrSF` * `HasPool` = `PoolArea` > 0 * `HasGarage` = `GarageArea` > 0 * `HasFireplace` = `Fireplaces` > 0 #### Encoding Categorical Features I used one-hot encoding to convert categorical features into numerical ones. ### Model Training and Evaluation I trained several regression models (Linear Regression, Ridge Regression, Lasso Regression, ElasticNet Regression) and evaluated them using cross-validation. The best performing model was Ridge Regression with an RMSE of around 0.12 on the validation set. ### Submission I made predictions on the test set using the best performing model and submitted them to Kaggle. ## Results My best submission achieved an RMSE of around 0.12 on the public leaderboard. <|repo_name|>zzangpp/kaggle-competitions<|file_sep|>/house-prices-advanced-regression-techniques/data_analysis.py import pandas as pd import numpy as np import matplotlib.pyplot as plt import seaborn as sns def load_data(): train_df = pd.read_csv('train.csv') test_df = pd.read_csv('test.csv') return train_df, test_df def plot_correlation_matrix(df): corr = df.corr() plt.figure(figsize=(10,10)) sns.set(font_scale=1) sns.heatmap(corr, cmap='RdBu', vmin=-1, vmax=1, annot=True) def plot_feature_scatter(df): cols = ['SalePrice', 'OverallQual', 'GrLivArea', 'GarageCars', 'TotalBsmtSF', 'FullBath', 'YearBuilt'] # Scatter plot #sns.set() #sns.pairplot(df[cols], size = 2) #plt.show() # New figure fig = plt.figure(figsize=(20,15)) for i in range(1,len(cols)): ax = fig.add_subplot(3,int(len(cols)/3)+1,i) ax.scatter(df[cols[0]], df[cols[i]], marker='o') ax.set_xlabel(cols[i]) ax.set_ylabel(cols[0]) plt.show() def plot_feature_hist(df): cols = ['SalePrice', 'OverallQual', 'GrLivArea', 'GarageCars', 'TotalBsmtSF', 'FullBath', 'YearBuilt'] # Histogram plot #sns.set() #sns.pairplot(df[cols], size = 2) #plt.show() # New figure fig = plt.figure(figsize=(20,15)) for i in range(1,len(cols)): ax = fig.add_subplot(3,int(len(cols)/3)+1,i) ax.hist(df[df[cols[i]] != np.nan][cols[i]], bins=20) ax.set_xlabel(cols[i]) ax.set_ylabel('Count') plt.show() def plot_missing_data(df): total = df.isnull().sum().sort_values(ascending=False) percent = (df.isnull().sum()/df.isnull().count()).sort_values(ascending=False) missing_data = pd.concat([total, percent], axis=1, keys=['Total', 'Percent']) print(missing_data.head(20)) def main(): train_df, test_df = load_data() combined_df = pd.concat([train_df.drop(['SalePrice'], axis=1), test_df]) plot_correlation_matrix(combined_df) plot_feature_scatter(train_df) plot_feature_hist(train_df) plot_missing_data(combined_df) if __name__ == '__main__': main() <|repo_name|>zzangpp/kaggle-competitions<|file_sep|>/house-prices-advanced-regression-techniques/feature_engineering.py import pandas as pd import numpy as np def load_data(): train_df = pd.read_csv('train.csv') test_df = pd.read_csv('test.csv') return train_df, test_df def fill_missing_values(df): numeric_cols_with_missing_values = df.select_dtypes(include=[np.number]).columns[df.isnull().any()].tolist() for col in numeric_cols_with_missing_values: df[col] = df[col].fillna(df[col].median()) categorical_cols_with_missing_values = df.select_dtypes(include=[object]).columns[df.isnull().any()].tolist() for col in categorical_cols_with_missing_values: df[col] = df[col].fillna(df[col].mode()[0]) return df def add_features(df): # Total SF df['TotalSF'] = df['TotalBsmtSF'] + df['1stFlrSF'] + df['2ndFlrSF'] # Has Pool? df['HasPool'] = (df['PoolArea'] > 0).astype(int) # Has Garage? df['HasGarage'] = (df['GarageArea'] > 0).astype(int) # Has Fireplace? df['HasFireplace'] = (df['Fireplaces'] > 0).astype(int) return df def one_hot_encode(df): categorical_cols_to_encode = df.select_dtypes(include=[object]).columns.tolist() return pd.get_dummies(df[categorical_cols_to_encode]) def main(): train_df, test_df = load_data() train_df_filled_in_missing_values = fill_missing_values(train_df) test_df_filled_in_missing_values = fill_missing_values(test_df) train_df_features_added = add_features(train_df_filled_in_missing_values) test_df_features_added = add_features(test_df_filled_in_missing_values) train_features_encoded_one_hot_encoded_features_added_and_filled_in_missing_values = one_hot_encode(train_df_features_added).join(train_features_added[['Id', 'SalePrice']]) test_features_encoded_one_hot_encoded_features_added_and_filled_in_missing_values = one_hot_encode(test_df_features_added).join(test_features_added[['Id']]) if __name__ == '__main__': main()<|repo_name|>zzangpp/kaggle-competitions<|file_sep|>/house-prices-advanced-regression-techniques/model_training.py import pandas as pd import numpy as np from sklearn.model_selection import cross_val_score from sklearn.linear_model import LinearRegression, RidgeCV,LassoCV,ElasticNetCV def load_data(): train_labels_filename_pathname_stringy_stringy_stringy_stringy_stringy_stringy_stringy_stringy ='train_labels_encoded_one_hot_encoded_features_added_and_filled_in_missing_values.csv'