# Data setup
import warnings; warnings.filterwarnings("ignore")
import pandas as pd
import seaborn as sns
import numpy as np
np.random.seed(42)
tips = sns.load_dataset("tips")  # Load sample tips dataset
print("Shape of the dataset:", tips.shape)
print("")
print("First few rows:")
print(tips.head())

# Basic data summary
print("Columns:", tips.columns.tolist())
print("")
print("Missing values in each column:")
print(tips.isnull().sum())

# Remove rows with missing values just in case
tips = tips.dropna()
print("New data shape after dropping missing values:", tips.shape)

# Visualize the distribution of tips by sex
import matplotlib.pyplot as plt
sns.histplot(data=tips, x="tip", hue="sex", element="step", stat="density", common_norm=False, bins=20, palette="muted")
plt.title("Tip Amounts by Sex")
plt.xlabel("Tip Amount (USD)")
plt.ylabel("Density")
plt.legend(title="Sex")
plt.show()

# Quick look at average tip for each group
avg_tips = tips.groupby("sex")["tip"].mean()
print("Average tip by sex:")
print(avg_tips)

# Two-sample t-test for difference in means
from scipy.stats import ttest_ind
male_tip = tips.loc[tips.sex == "Male", "tip"]
female_tip = tips.loc[tips.sex == "Female", "tip"]
t_stat, p_value = ttest_ind(male_tip, female_tip)
print("T-statistic:", t_stat)
print("P-value:", p_value)

# Interpret results
alpha = 0.05
if p_value < alpha:
    print("We reject the null hypothesis. There is evidence that males and females tip differently.")
else:
    print("We cannot reject the null hypothesis. There is not enough evidence of a difference.")
    

# Practice: Try it yourself!
col = input("Choose a column to compare by (sex, smoker, time, day): ")
grp1 = input("Enter the first group name as shown in the dataset (example: Male): ")
grp2 = input("Enter the second group name: ")

data1 = tips.loc[tips[col] == grp1, "tip"]
data2 = tips.loc[tips[col] == grp2, "tip"]

t_stat2, p2 = ttest_ind(data1, data2)
print(f"Comparing average tips for {grp1} versus {grp2}")
print("t-statistic:", t_stat2)
print("p-value:", p2)
if p2 < 0.05:
    print("Result: There is a significant difference!")
else:
    print("Result: There is no significant difference.")
    

# Mini project: Test if day of week affects tipping
for d in tips["day"].unique():
    tip_for_day = tips.loc[tips["day"] == d, "tip"]
    print(f"{d}: mean tip = {tip_for_day.mean():.2f}, count = {tip_for_day.count()}")
    

# ANOVA: Are tips different across days?
from scipy.stats import f_oneway
day_tips = [tips.loc[tips["day"] == d, "tip"] for d in tips["day"].unique()]
f_stat, pval_day = f_oneway(*day_tips)
print("F-statistic:", f_stat)
print("P-value:", pval_day)

# Boxplot of tips by day to visualize results
plt.figure(figsize=(8,5))
sns.boxplot(data=tips, x="day", y="tip", palette="Set3")
plt.title("Distribution of Tips by Day")
plt.xlabel("Day of Week")
plt.ylabel("Tip (USD)")
plt.show()

# Troubleshooting: Common mistakes
try:
    nonexist = tips.loc[tips['bad_col'] == 'broken', 'tip']
except Exception as e:
    print("Error: column not found. Always check spelling and column names.")
    print("More info:", e)