# Importing all the libraries we will need
import warnings; warnings.filterwarnings('ignore')
import numpy as np
import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt

# Data setup
# We will use the Tips dataset for our examples
tips = sns.load_dataset('tips')
print('Dataset shape:', tips.shape)
tips.head()

# Simulate flipping a coin 10 times
np.random.seed(42)
flips = np.random.choice(['heads', 'tails'], size=10)
print('Results:', flips)
heads_count = np.sum(flips == 'heads')
print('Number of heads:', heads_count)

# Calculate the probability of getting heads in our simulation
prob_heads = heads_count / len(flips)
print('Simulated probability of heads:', prob_heads)

# Calculate mean and median tip
mean_tip = tips['tip'].mean()
median_tip = tips['tip'].median()
print('Mean tip:', mean_tip)
print('Median tip:', median_tip)

# Check for missing data
missing = tips.isnull().sum()
print('Missing values in each column:')
print(missing)

# Remove rows with missing values
tips_clean = tips.dropna()
print('Shape after removing missing data:', tips_clean.shape)

# Visualize tip amounts by gender
sns.boxplot(x='sex', y='tip', data=tips_clean)
plt.title('Tip Amounts by Gender')
plt.show()

# State the hypothesis
print('Null hypothesis: Average tip is the same for men and women.')
print('Alternative hypothesis: Average tip is different for men and women.')

# Observed mean difference in tip by gender
mean_male = tips_clean[tips_clean['sex'] == 'Male']['tip'].mean()
mean_female = tips_clean[tips_clean['sex'] == 'Female']['tip'].mean()
obs_diff = mean_male - mean_female
print('Observed difference in mean tip (Male - Female):', obs_diff)

# Run a permutation test
n_permutations = 1000
combined = tips_clean['tip'].values
labels = tips_clean['sex'].values
diffs = []
for _ in range(n_permutations):
    shuffled_labels = np.random.permutation(labels)
    mean_male = combined[shuffled_labels == 'Male'].mean()
    mean_female = combined[shuffled_labels == 'Female'].mean()
    diffs.append(mean_male - mean_female)
diffs = np.array(diffs)

# Plot permutation differences and observed value
plt.hist(diffs, bins=30, color='skyblue', edgecolor='black')
plt.axvline(obs_diff, color='red', linestyle='--', label='Observed diff')
plt.xlabel('Mean difference (Male - Female)')
plt.ylabel('Count')
plt.title('Permutation Test Results')
plt.legend()
plt.show()

# Calculate the p-value
p_value = np.mean(np.abs(diffs) >= np.abs(obs_diff))
print('Estimated p-value:', p_value)

# Try a two-group comparison for smokers vs non-smokers
mean_smoker = tips_clean[tips_clean['smoker'] == 'Yes']['tip'].mean()
mean_nonsmoker = tips_clean[tips_clean['smoker'] == 'No']['tip'].mean()
obs_diff_smoker = mean_smoker - mean_nonsmoker
print('Observed difference in mean tip (Smoker - Non-Smoker):', obs_diff_smoker)