Mini Project: Titanic Survival Classifier

import numpy as np
import pandas as pd
from sklearn.model_selection import train_test_split, cross_val_score, GridSearchCV
from sklearn.compose import ColumnTransformer
from sklearn.preprocessing import StandardScaler, OneHotEncoder, OrdinalEncoder
from sklearn.impute import SimpleImputer
from sklearn.pipeline import Pipeline
from sklearn.linear_model import LogisticRegression
from sklearn.ensemble import RandomForestClassifier, GradientBoostingClassifier
from sklearn.metrics import classification_report, roc_auc_score, confusion_matrix
import matplotlib.pyplot as plt
import seaborn as sns

# SIMULATE TITANIC DATA (realistic proportions)
np.random.seed(42)
N = 891
df = pd.DataFrame({
    "Pclass":    np.random.choice([1, 2, 3], N, p=[0.25, 0.21, 0.54]),
    "Sex":       np.random.choice(["male", "female"], N, p=[0.65, 0.35]),
    "Age":       np.random.normal(29, 14, N).clip(0.5, 80),
    "SibSp":     np.random.choice(range(9), N, p=[0.68, 0.23, 0.03, 0.02, 0.01, 0.01, 0.01, 0.005, 0.005]),
    "Parch":     np.random.choice(range(7), N, p=[0.76, 0.13, 0.09, 0.01, 0.005, 0.003, 0.002]),
    "Fare":      np.random.exponential(32, N).clip(0, 512),
    "Embarked":  np.random.choice(["S", "C", "Q", None], N, p=[0.72, 0.19, 0.09, 0.002]),
})
# Survival: women and higher class had better survival (historical)
df["Survived"] = (
    (df["Sex"] == "female").astype(int) * 0.5 +
    (df["Pclass"] == 1).astype(int) * 0.3 +
    np.random.uniform(0, 1, N) * 0.2
) > 0.4
df["Survived"] = df["Survived"].astype(int)

# Inject missing values (realistic)
df.loc[df.sample(177, random_state=1).index, "Age"]      = np.nan
df.loc[df.sample(2, random_state=2).index, "Embarked"]  = None
print(f"Dataset: {df.shape} | Survival rate: {df['Survived'].mean():.1%}")

# FEATURE ENGINEERING
df["FamilySize"] = df["SibSp"] + df["Parch"] + 1
df["IsAlone"]    = (df["FamilySize"] == 1).astype(int)
df["ChildGuard"] = ((df["Age"] < 16) & (df["Parch"] > 0)).astype(int)

feature_cols = ["Pclass", "Sex", "Age", "Fare", "Embarked", "FamilySize", "IsAlone", "ChildGuard"]
X = df[feature_cols]
y = df["Survived"]

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42, stratify=y)

# PREPROCESSING
num_cols = ["Age", "Fare", "FamilySize", "IsAlone", "ChildGuard"]
cat_cols  = ["Sex", "Embarked"]
ord_cols  = ["Pclass"]

preprocessor = ColumnTransformer([
    ("num", Pipeline([("imp", SimpleImputer(strategy="median")), ("sc", StandardScaler())]), num_cols),
    ("cat", Pipeline([("imp", SimpleImputer(strategy="most_frequent")), ("ohe", OneHotEncoder(drop="first", sparse_output=False, handle_unknown="ignore"))]), cat_cols),
    ("ord", OrdinalEncoder(categories=[[1, 2, 3]]), ord_cols),
])

# MODEL COMPARISON
models = {
    "Logistic Regression": LogisticRegression(C=0.5, class_weight="balanced", max_iter=1000, random_state=42),
    "Random Forest":       RandomForestClassifier(n_estimators=100, max_depth=6, class_weight="balanced", random_state=42),
    "Gradient Boosting":   GradientBoostingClassifier(n_estimators=200, learning_rate=0.05, max_depth=3, random_state=42),
}

print("\nCV AUC-ROC (5-fold):")
for name, model in models.items():
    pipe = Pipeline([("prep", preprocessor), ("model", model)])
    auc = cross_val_score(pipe, X_train, y_train, cv=5, scoring="roc_auc").mean()
    print(f"  {name:<22}: {auc:.4f}")

# BEST MODEL: GRADIENT BOOSTING
best_pipe = Pipeline([("prep", preprocessor), ("model", GradientBoostingClassifier(n_estimators=200, learning_rate=0.05, max_depth=3, random_state=42))])
best_pipe.fit(X_train, y_train)

y_pred  = best_pipe.predict(X_test)
y_prob  = best_pipe.predict_proba(X_test)[:, 1]
test_auc = roc_auc_score(y_test, y_prob)

print(f"\nTest AUC-ROC: {test_auc:.4f}")
print(classification_report(y_test, y_pred, target_names=["Did not survive", "Survived"]))

# PREDICTION FUNCTION
def predict_survival(pclass: int, sex: str, age: float, fare: float) -> str:
    sample = pd.DataFrame([{"Pclass": pclass, "Sex": sex, "Age": age, "Fare": fare,
                             "Embarked": "S", "FamilySize": 1, "IsAlone": 1, "ChildGuard": 0}])
    prob = best_pipe.predict_proba(sample)[0, 1]
    return f"{prob:.1%} survival probability"

print("\nPredictions:")
print(f"  1st class female, 28: {predict_survival(1, 'female', 28, 100)}")
print(f"  3rd class male, 35:   {predict_survival(3, 'male', 35, 8)}")