import numpy as np
import matplotlib.pyplot as plt

# Tommy's data
temperatures = [70, 85, 75, 90]  # °F
cups_sold = [10, 23, 14, 28]

# Calculate the relationship (simplified linear regression)
avg_temp = np.mean(temperatures)
avg_cups = np.mean(cups_sold)

# Find the slope (how much sales change per degree)
numerator = sum((t - avg_temp) * (c - avg_cups) 
                for t, c in zip(temperatures, cups_sold))
denominator = sum((t - avg_temp) ** 2 for t in temperatures)
slope = numerator / denominator

# Find the intercept (base sales)
intercept = avg_cups - slope * avg_temp

print(f"🎯 Tommy's Magic Formula:")
print(f"Cups = {intercept:.2f} + {slope:.2f} × Temperature")
print(f"\nFor every 1°F increase, Tommy sells {slope:.2f} more cups!")

# Predict for 80°F
prediction = intercept + slope * 80
print(f"\n📊 Prediction for 80°F: {prediction:.1f} cups")

import pandas as pd
import numpy as np
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LinearRegression
from sklearn.metrics import mean_absolute_error, r2_score

# Load MegaMart data (simulated)
print("📊 Loading MegaMart Sales Data...")
print("="*50)

# Create features that capture business patterns
features = {
    'temperature': 75,           # Weather data
    'day_of_week': 'Monday',    # Weekly patterns
    'is_weekend': 0,             # Weekend flag
    'is_campaign': 1,            # Marketing campaign active
    'competitor_sale': 0,        # Competitor having sale
    'social_buzz': 150,          # Social media mentions
    'avg_price': 249,            # Average product price
    'days_to_holiday': 45,       # Days until next major holiday
}

print("🔧 Engineering Features:")
print("-"*50)

# Cyclical encoding for time features
features['day_sin'] = np.sin(2 * np.pi * 1 / 7)  # Monday = 1
features['day_cos'] = np.cos(2 * np.pi * 1 / 7)
features['month_sin'] = np.sin(2 * np.pi * 10 / 12)  # October
features['month_cos'] = np.cos(2 * np.pi * 10 / 12)

# Interaction features
features['weekend_campaign'] = features['is_weekend'] * features['is_campaign']
features['price_sensitivity'] = features['avg_price'] * features['competitor_sale']

# Display engineered features
for feature, value in features.items():
    print(f"  • {feature}: {value}")

print("\n✨ Total Features: ", len(features))
print("🎯 Ready to train the model!")

# Train the model
print("🤖 TRAINING LINEAR REGRESSION MODEL")
print("="*50)

# Simulated training process
X_train_size = 292
X_test_size = 73

print(f"📚 Training set: {X_train_size} days of data")
print(f"🧪 Test set: {X_test_size} days of data")
print("\n⏳ Training in progress...")

# Simulate training (in real scenario, this would be actual sklearn code)
import time
for i in range(3):
    print(f"   Iteration {i+1}/3... ", end="")
    time.sleep(0.5)
    print("✓")

print("\n✅ Training complete!")
print("\n📈 MODEL PERFORMANCE:")
print("="*50)

# Results
train_mae = 487.23
test_mae = 512.45
train_r2 = 0.891
test_r2 = 0.878

print(f"Training Set:")
print(f"  MAE: ${train_mae:,.2f}")
print(f"  R² Score: {train_r2:.3f}")
print(f"  Accuracy: {100 - (train_mae/10000*100):.1f}%")

print(f"\nTest Set (Unseen Data):")
print(f"  MAE: ${test_mae:,.2f}")
print(f"  R² Score: {test_r2:.3f}")
print(f"  Accuracy: {100 - (test_mae/10000*100):.1f}%")

print("\n🔍 TOP 5 MOST IMPORTANT FEATURES:")
print("="*50)
important_features = [
    ('sales_lag_1', 312.45, 'Yesterday\'s momentum'),
    ('is_weekend', 287.23, 'Weekend shopping surge'),
    ('temperature', 198.67, 'Weather impact'),
    ('social_buzz', 156.89, 'Viral product trends'),
    ('days_to_holiday', 142.34, 'Holiday shopping urgency')
]

for feature, impact, description in important_features:
    print(f"  {feature}: +${impact:.2f} per unit - {description}")