Fixes : #6551

machine_learning/xgboostclassifier.py

@@ -0,0 +1,79 @@
+import matplotlib.pyplot as plt
+
+# https://xgboost.readthedocs.io/en/stable/
+import numpy as np
+import pandas as pd
+import seaborn as sns
+from xgboost import XGBClassifier
+
+training = pd.read_csv("../input/titanic/train.csv")
+test = pd.read_csv("../input/titanic/test.csv")
+
+# Commented out IPython magic to ensure Python compatibility.
+training["train_test"] = 1
+test["train_test"] = 0
+test["Survived"] = np.NaN
+all_data = pd.concat([training, test])
+# %matplotlib inline
+all_data.columns
+
+all_data.describe()
+
+all_data["cabin_mul"] = all_data.Cabin.apply(
+    lambda x: 0 if pd.isna(x) else len(x.split(" "))
+)
+all_data["cabin_adv"] = all_data.Cabin.apply(lambda x: str(x)[0])
+all_data["name_title"] = all_data.Name.apply(
+    lambda x: x.split(",")[1].split(".")[0].strip()
+)
+all_data.Age = all_data.Age.fillna(training.Age.median())
+all_data.Fare = all_data.Fare.fillna(training.Fare.median())
+all_data.dropna(subset=["Embarked"], inplace=True)
+all_data["norm_fare"] = np.log(all_data.Fare + 1)
+all_data.Pclass = all_data.Pclass.astype(str)
+all_data["Age"] = all_data["Age"].apply(np.int64)
+all_dummies = pd.get_dummies(
+    all_data[
+        [
+            "Pclass",
+            "Sex",
+            "Age",
+            "SibSp",
+            "Parch",
+            "norm_fare",
+            "Embarked",
+            "cabin_adv",
+            "cabin_mul",
+            "name_title",
+            "train_test",
+        ]
+    ]
+)
+
+from sklearn.preprocessing import StandardScaler
+
+scale = StandardScaler()
+all_dummies_scaled = all_dummies.copy()
+all_dummies_scaled[["Age", "SibSp", "Parch", "norm_fare"]] = scale.fit_transform(
+    all_dummies_scaled[["Age", "SibSp", "Parch", "norm_fare"]]
+)
+all_dummies_scaled.head()
+
+X_train_scaled = all_dummies_scaled[all_dummies_scaled.train_test == 1].drop(
+    ["train_test"], axis=1
+)
+X_test_scaled = all_dummies_scaled[all_dummies_scaled.train_test == 0].drop(
+    ["train_test"], axis=1
+)
+
+y_train = all_data[all_data.train_test == 1].Survived
+
+from xgboost import XGBClassifier
+
+xgb = XGBClassifier()
+xgb.fit(X_train_scaled, y_train)
+
+y_hat_base_vc = xgb.predict(X_test_scaled).astype(int)
+basic_submission = {"PassengerId": test.PassengerId, "Survived": y_hat_base_vc}
+base_submission = pd.DataFrame(data=basic_submission)
+base_submission.to_csv("xgb_submission.csv", index=False)

machine_learning/xgboostregressor.py

@@ -0,0 +1,164 @@
+import os
+
+import matplotlib.pyplot as plt
+
+# https://xgboost.readthedocs.io/en/stable/
+import numpy as np
+import pandas as pd
+import seaborn as sns
+from xgboost import XGBRegressor
+
+for dirname, _, filenames in os.walk("/kaggle/input"):
+    for filename in filenames:
+        print(os.path.join(dirname, filename))
+
+trainAmes = pd.read_csv("/kaggle/input/ames-housing-dataset/AmesHousing.csv")
+test = pd.read_csv("/kaggle/input/house-prices-advanced-regression-techniques/test.csv")
+train = pd.read_csv(
+    "/kaggle/input/house-prices-advanced-regression-techniques/train.csv"
+)
+
+trainAmes.columns = trainAmes.columns.str.replace(" ", "")
+trainAmes = trainAmes.rename(columns={"YearRemod/Add": "YearRemodAdd"})
+
+data = pd.concat([trainAmes, train, test], axis=0, sort=False)
+print("Size of the Housing Dataset", len(data))
+useless = ["Id", "PID", "Order", "SalePrice"]
+data = data.drop(useless, axis=1)
+duplicate = data[data.duplicated(keep="last")].index
+len(duplicate)
+
+duplicate = duplicate[0:390]
+trainAmes = trainAmes.drop(duplicate, axis=0)
+
+training = pd.concat([trainAmes, train], axis=0, sort=False)
+useless = ["Id", "PID", "Order"]
+training = training.drop(useless, axis=1)
+
+# Separating Target and Features
+
+target = training["SalePrice"]
+test_id = test["Id"]
+test = test.drop(["Id"], axis=1)
+training2 = training.drop(["SalePrice"], axis=1)
+
+
+# Concatenating train & test set
+
+train_test = pd.concat([training2, test], axis=0, sort=False)
+
+# Filling Categorical NaN (That we know how to fill due to the description file )
+
+train_test["Functional"] = train_test["Functional"].fillna("Typ")
+train_test["Electrical"] = train_test["Electrical"].fillna("SBrkr")
+train_test["KitchenQual"] = train_test["KitchenQual"].fillna("TA")
+train_test["Exterior1st"] = train_test["Exterior1st"].fillna(
+    train_test["Exterior1st"].mode()[0]
+)
+train_test["Exterior2nd"] = train_test["Exterior2nd"].fillna(
+    train_test["Exterior2nd"].mode()[0]
+)
+train_test["SaleType"] = train_test["SaleType"].fillna(train_test["SaleType"].mode()[0])
+train_test["PoolQC"] = train_test["PoolQC"].fillna("None")
+train_test["Alley"] = train_test["Alley"].fillna("None")
+train_test["FireplaceQu"] = train_test["FireplaceQu"].fillna("None")
+train_test["Fence"] = train_test["Fence"].fillna("None")
+train_test["MiscFeature"] = train_test["MiscFeature"].fillna("None")
+for col in ("GarageArea", "GarageCars"):
+    train_test[col] = train_test[col].fillna(0)
+
+for col in ["GarageType", "GarageFinish", "GarageQual", "GarageCond"]:
+    train_test[col] = train_test[col].fillna("None")
+
+for col in ("BsmtQual", "BsmtCond", "BsmtExposure", "BsmtFinType1", "BsmtFinType2"):
+    train_test[col] = train_test[col].fillna("None")
+
+for col in (
+    "BsmtFinSF1",
+    "BsmtFinSF2",
+    "BsmtFullBath",
+    "BsmtHalfBath",
+    "MasVnrArea",
+    "BsmtUnfSF",
+    "TotalBsmtSF",
+):
+    train_test[col] = train_test[col].fillna(0)
+
+train_test["LotFrontage"] = train_test["LotFrontage"].fillna(
+    train["LotFrontage"].median()
+)
+
+# Checking the features with NaN remained out
+
+for col in train_test:
+    if train_test[col].isna().sum() > 0:
+        print(train_test[col][1])
+
+# Converting non-numeric predictors stored as numbers into string
+
+train_test["MSSubClass"] = train_test["MSSubClass"].apply(str)
+train_test["YrSold"] = train_test["YrSold"].apply(str)
+train_test["MoSold"] = train_test["MoSold"].apply(str)
+train_test["OverallQual"] = train_test["OverallQual"].apply(str)
+train_test["OverallCond"] = train_test["OverallCond"].apply(str)
+train_test["SqFtPerRoom"] = train_test["GrLivArea"] / (
+    train_test["TotRmsAbvGrd"]
+    + train_test["FullBath"]
+    + train_test["HalfBath"]
+    + train_test["KitchenAbvGr"]
+)
+
+train_test["Total_Home_Quality"] = train_test["OverallQual"] + train_test["OverallCond"]
+
+train_test["Total_Bathrooms"] = (
+    train_test["FullBath"]
+    + (0.5 * train_test["HalfBath"])
+    + train_test["BsmtFullBath"]
+    + (0.5 * train_test["BsmtHalfBath"])
+)
+
+train_test["HighQualSF"] = train_test["1stFlrSF"] + train_test["2ndFlrSF"]
+train_test["renovated"] = train_test["YearRemodAdd"] + train_test["YearBuilt"]
+
+# Removing the useless variables
+
+useless = ["GarageYrBlt", "YearRemodAdd"]
+train_test = train_test.drop(useless, axis=1)
+# Creating dummy variables from categorical features
+
+from scipy.stats import skew
+
+train_test_dummy = pd.get_dummies(train_test)
+
+numeric_features = train_test_dummy.dtypes[train_test_dummy.dtypes != object].index
+skewed_features = (
+    train_test_dummy[numeric_features]
+    .apply(lambda x: skew(x))
+    .sort_values(ascending=False)
+)
+high_skew = skewed_features[skewed_features > 0.5]
+skew_index = high_skew.index
+
+# Normalize skewed features using log_transformation
+
+for i in skew_index:
+    train_test_dummy[i] = np.log1p(train_test_dummy[i])
+
+target_log = np.log1p(target)
+
+from xgboost import XGBRegressor
+
+# Train-Test separation
+
+X_train = train_test_dummy[0:4000]
+X_test = train_test_dummy[4000:]
+
+xgb = XGBRegressor()
+xgb.fit(X_train, target_log)
+
+test_pred = xgb.predict(X_test)
+submission = pd.DataFrame(test_id, columns=["Id"])
+test_pred = np.expm1(test_pred)
+submission["SalePrice"] = test_pred
+submission.head()
+submission.to_csv("xgb.csv", index=False, header=True)