import warnings
warnings.filterwarnings("ignore")

# We will often use pandas and matplotlib for data tasks
import pandas as pd
import matplotlib.pyplot as plt

# Data setup: Let's load a real dataset about airline passengers over time
url = "https://raw.githubusercontent.com/jbrownlee/Datasets/master/airline-passengers.csv"
data = pd.read_csv(url)
print("Data shape:", data.shape)
data.head()

# Plot the passenger numbers over time
plt.figure(figsize=(10,4))
plt.plot(data['Month'], data['Passengers'], label="Passengers")
plt.xlabel("Month")
plt.ylabel("Number of Passengers")
plt.title("Monthly Airline Passengers")
plt.xticks(rotation=45)
plt.legend()
plt.tight_layout()
plt.show()

# Let us build two small lists and see how covariance works
score_math = [55, 85, 67, 90, 77]
score_science = [60, 80, 70, 95, 79]

scores = pd.DataFrame({"Math": score_math, "Science": score_science})
print(scores)

print("Covariance matrix:")
print(scores.cov())

# Correlation works very much like covariance but gives numbers between -1 and +1.
print("Correlation matrix:")
print(scores.corr())

# Try changing the scale. Multiply math scores by 10.
scores["Math_x10"] = scores["Math"] * 10

print(scores[["Math_x10", "Science"]].cov())
print(scores[["Math_x10", "Science"]].corr())

# Correlation in a real-world dataset (airline passengers)
corr = data['Passengers'].corr(pd.Series(range(len(data))))
print(f"Correlation between 'Passengers' and months: {corr:.2f}")

# Compare two years: are 1958 and 1959 related?
data["Year"] = data["Month"].str[:4].astype(int)
yearly = data.groupby("Year")["Passengers"].sum()
print(yearly)
print("Covariance and correlation:")
print(yearly.cov(yearly.shift(1)))
print(yearly.corr(yearly.shift(1)))

# Missing data can sometimes happen in the real world
data_missing = data.copy()
data_missing.loc[0, "Passengers"] = None
data_missing.loc[1, "Passengers"] = None
print("Does pandas handle missing numbers?")
print(data_missing["Passengers"].corr(pd.Series(range(len(data_missing)))))

# How to deal with missing data: fill or drop?
# Option 1: Fill missing with the average value
filled = data_missing["Passengers"].fillna(data_missing["Passengers"].mean())
print(filled.corr(pd.Series(range(len(filled)))))

# Option 2: Remove all missing rows
dropped = data_missing.dropna()
print(dropped["Passengers"].corr(pd.Series(range(len(dropped)))))

# Using input() to enter two lists and compare their correlation
user1 = input("List one, comma separated numbers: ")
user2 = input("List two, comma separated numbers: ")
lst1 = [float(x.strip()) for x in user1.split(",")]
lst2 = [float(x.strip()) for x in user2.split(",")]
print("Covariance:", pd.Series(lst1).cov(pd.Series(lst2)))
print("Correlation:", pd.Series(lst1).corr(pd.Series(lst2)))

# Quick: Find the column pair in the airline data with the strongest correlation
numeric_data = data.select_dtypes(include='number')
print(numeric_data.corr())

# Creating a scatter plot to see the relationship
plt.scatter(scores["Math"], scores["Science"], color='red')
plt.xlabel("Math scores")
plt.ylabel("Science scores")
plt.title("Scatter Plot: Math vs. Science")
plt.show()

# Easy tip: Get correlation fast with numpy
import numpy as np
np.corrcoef(score_math, score_science)

# Mini project part 1: Shampoo sales vs. time
url2 = "https://raw.githubusercontent.com/jbrownlee/Datasets/master/shampoo.csv"
shampoo = pd.read_csv(url2)
shampoo.columns = ["Month", "Sales"]
shampoo["Sales"] = pd.to_numeric(shampoo["Sales"], errors="coerce")

print(shampoo.head())
plt.plot(shampoo['Sales'])
plt.title("Shampoo Sales Over Time")
plt.xlabel("Month")
plt.ylabel("Sales")
plt.show()

# Mini project part 2: Correlation between airline and shampoo trends
cor_air_shamp = pd.Series(data["Passengers"][:len(shampoo)]).corr(shampoo["Sales"])
print(f"Correlation between airline passenger and shampoo sales (first {len(shampoo)} rows): {cor_air_shamp:.2f}")

# Challenge: Can you find a pair with negative correlation?
lst_a = [10, 8, 6, 4, 2]
lst_b = [1, 3, 5, 7, 9]
print("Correlation:", pd.Series(lst_a).corr(pd.Series(lst_b)))