# Import libraries and suppress warnings
import warnings; warnings.filterwarnings('ignore')
import numpy as np
import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt

# Data setup: Load the Titanic dataset (to show spread on real data)
titanic = sns.load_dataset('titanic')
print('Titanic data shape:', titanic.shape)
titanic.head()

# Let us look at 'age' column: drop NA for simplicity
ages = titanic['age'].dropna()
print('Number of age values:', len(ages))
ages.head()

# Visualizing age distribution
plt.figure(figsize=(8,4))
sns.histplot(ages, bins=30, kde=True, color='blue')
plt.title('Age Distribution on Titanic')
plt.xlabel('Age')
plt.ylabel('Count')
plt.show()

# Demo: variance step by step for ages 10, 22, 30, 40, 50
demo_ages = [10, 22, 30, 40, 50]
mean_demo = np.mean(demo_ages)
print('Mean age:', mean_demo)
squared_diffs = [(x - mean_demo) ** 2 for x in demo_ages]
print('Squared differences:', squared_diffs)
variance_demo = np.mean(squared_diffs)
print('Variance:', variance_demo)

# Standard deviation is just the square root of variance
std_demo = np.sqrt(variance_demo)
print('Standard deviation:', std_demo)

# Calculating with pandas for real Titanic data
titanic_age_var = ages.var(ddof=0)
titanic_age_std = ages.std(ddof=0)
print('Titanic Age Variance:', titanic_age_var)
print('Titanic Age Standard Deviation:', titanic_age_std)

# Compare two groups: Survivors vs Non-survivors' age spread
survived_ages = titanic[titanic['survived'] == 1]['age'].dropna()
not_survived_ages = titanic[titanic['survived'] == 0]['age'].dropna()
print('Survivors std:', survived_ages.std(ddof=0))
print('Non-survivors std:', not_survived_ages.std(ddof=0))

# Visualize both age distributions overlaid
plt.figure(figsize=(8,4))
sns.histplot(survived_ages, bins=30, color='green', kde=True, label='Survived', alpha=0.5)
sns.histplot(not_survived_ages, bins=30, color='red', kde=True, label='Did Not Survive', alpha=0.5)
plt.legend()
plt.title('Age Spread: Survived vs Not Survived')
plt.xlabel('Age')
plt.ylabel('Count')
plt.show()

# Calculate IQR by hand and with pandas
q1 = ages.quantile(0.25)
q3 = ages.quantile(0.75)
iqr = q3 - q1
print('25th percentile (Q1):', q1)
print('75th percentile (Q3):', q3)
print('Interquartile Range (IQR):', iqr)

# Visualize age spread with a boxplot to show IQR and outliers
plt.figure(figsize=(8,2))
sns.boxplot(x=ages, color='skyblue')
plt.title('Boxplot of Ages on Titanic')
plt.xlabel('Age')
plt.show()

# Compare spread with and without outliers: add some fake ages
ages_extended = pd.concat([ages, pd.Series([5, 7, 8, 99, 100, 101])], ignore_index=True)
print('Original std:', ages.std(ddof=0))
print('With outliers std:', ages_extended.std(ddof=0))
print('Original IQR:', ages.quantile(0.75) - ages.quantile(0.25))
print('With outliers IQR:', ages_extended.quantile(0.75) - ages_extended.quantile(0.25))

# Practice: Calculate variance, standard deviation, and IQR for 'fare' prices
fares = titanic['fare'].dropna()
print('Fare variance:', fares.var(ddof=0))
print('Fare std:', fares.std(ddof=0))
print('Fare IQR:', fares.quantile(0.75) - fares.quantile(0.25))

# Challenge: Enter three numbers and get std and IQR back
nums = []
for i in range(3):
    value = float(input("Enter any number: "))
    nums.append(value)
series = pd.Series(nums)
print('Standard deviation:', series.std(ddof=0))
print('Interquartile range:', series.quantile(0.75) - series.quantile(0.25))