# Let us start by making sure warnings do not fill up our notebook
import warnings
warnings.filterwarnings("ignore")

# Here is a basic Python variable assignment
tree = "A system for making yes or no decisions!"
print(tree)

# Data setup
import pandas as pd
url = "https://raw.githubusercontent.com/datasciencedojo/datasets/master/titanic.csv"
df = pd.read_csv(url)
print("Shape of data:", df.shape)
df.head()

# Let us see our target: who survived?
df["Survived"].value_counts()

# Let us split data into training and testing parts
from sklearn.model_selection import train_test_split
features = ["Pclass", "Sex", "Age", "Fare"]
# Convert text to numbers
df["Sex"] = df["Sex"].map({"male": 0, "female": 1})
df["Age"].fillna(df["Age"].median(), inplace=True)
X = df[features]
y = df["Survived"]
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
print("Train size:", X_train.shape, "Test size:", X_test.shape)

# Build and train a decision tree classifier
from sklearn.tree import DecisionTreeClassifier
tree = DecisionTreeClassifier(random_state=42)
tree.fit(X_train, y_train)
print("Model trained!")

# Use our trained tree to predict
predictions = tree.predict(X_test)
print("Predictions:", predictions[:10])

# How accurate is our tree?
from sklearn.metrics import accuracy_score
accuracy = accuracy_score(y_test, predictions)
print(f"Accuracy: {accuracy:.2f}")

# Build and train a random forest
from sklearn.ensemble import RandomForestClassifier
forest = RandomForestClassifier(n_estimators=100, random_state=42)
forest.fit(X_train, y_train)
forest_preds = forest.predict(X_test)
print("Random forest predictions:", forest_preds[:10])

# Test how well the forest performs
forest_acc = accuracy_score(y_test, forest_preds)
print(f"Random forest accuracy: {forest_acc:.2f}")

# Which features mattered most?
import matplotlib.pyplot as plt
import numpy as np
feature_importances = forest.feature_importances_
features = np.array(features)
plt.bar(features, feature_importances)
plt.title("Feature Importance in Random Forest")
plt.ylabel("Importance")
plt.xlabel("Feature")
plt.show()

# Use input to try your own prediction!
pclass = int(input("Enter passenger class (1, 2, or 3): "))
sex = input("Enter sex (male or female): ")
age = float(input("Enter age: "))
fare = float(input("Enter fare: "))
sex_val = 0 if sex == "male" else 1
your_data = [[pclass, sex_val, age, fare]]
surv_pred = forest.predict(your_data)
if surv_pred[0]==1:
    print("Prediction: survived")
else:
    print("Prediction: did not survive")
    

# Challenge: Try out-of-bag score for your forest
forest_oob = RandomForestClassifier(n_estimators=100, oob_score=True, random_state=42)
forest_oob.fit(X_train, y_train)
print(f"Out-of-bag score: {forest_oob.oob_score_:.2f}")

# Common problems: missing values
import numpy as np
has_null = np.any(df[features].isnull())
print("Any missing values?", has_null)

# Handy: see more model details with classification report
from sklearn.metrics import classification_report
print(classification_report(y_test, forest_preds))

# Best practice: set a random_state for reproducible results
model_a = DecisionTreeClassifier(random_state=123)
model_b = DecisionTreeClassifier(random_state=456)
model_a.fit(X_train, y_train)
model_b.fit(X_train, y_train)
pred_a = model_a.predict(X_test)
pred_b = model_b.predict(X_test)
print(pred_a[:5], pred_b[:5])