import warnings
import numpy as np
import pandas as pd
np.random.seed(42)
warnings.filterwarnings('ignore')

# Data setup (Airline Passenger Flights Dataset)
import seaborn as sns
df = sns.load_dataset('flights')
print(df.shape)
print(df.head(3))

# Preview the original DataFrame structure
df.head()

# Pivot to wide format: years become columns
df_wide = df.pivot(index='month', columns='year', values='passengers')
print(df_wide.shape)
df_wide.head()

# Use melt to turn the wide DataFrame back into long format
df_long = df_wide.reset_index().melt(id_vars='month', var_name='year', value_name='passengers')
print(df_long.shape)
df_long.head()

# Stack example: move columns into indexed rows
stacked = df_wide.stack()
print(stacked.shape)
stacked.head(12)

# Notice the MultiIndex structure after stacking
print(stacked.index.names)

# Unstack example: convert the MultiIndex Series back to DataFrame
unstacked = stacked.unstack()
print(unstacked.shape)
unstacked.head()

# Visualize: compare original and melted forms
import matplotlib.pyplot as plt
plt.figure(figsize=(8,4))
plt.plot(df['passengers'], label='Original/order')
plt.plot(df_long.sort_values(['year','month'])['passengers'].values, '--', label='Melted/sorted')
plt.title('Passengers Over Time (Original vs Melted)')
plt.legend()
plt.show()

# Challenge: creating a tidy table for easy aggregation
df_tidy = df.copy()
df_tidy['date'] = pd.to_datetime(df_tidy['year'].astype(str) + '-' + df_tidy['month'].astype(str) + '-01')
df_tidy = df_tidy.sort_values('date')
df_tidy.set_index('date', inplace=True)
df_tidy.head()

# Stack and unstack with multi-level columns: creating multi-index columns
multi = df.pivot_table(index='month', columns='year', values='passengers', aggfunc='sum')
multi.head()

# Stack across columns: get a long Series from the table with multi-level columns
long_series = multi.stack()
print(long_series.shape)
long_series.head(10)

# Unstack to return the Series to a DataFrame
back_to_df = long_series.unstack()
back_to_df.head()

# Troubleshooting: Common issues with stack/unstack/melt
try:
    df_broken = df_wide.reset_index().melt(id_vars='wrong_column')
except Exception as e:
    print('Error:', e)
    

# Mini-project: Reshape and aggregate flights data
monthly_avg = df.pivot_table(index='month', values='passengers', aggfunc='mean')
print(monthly_avg)
max_passenger_month = monthly_avg['passengers'].idxmax()
print('Month with highest average passengers:', max_passenger_month)

# Your turn! Try reshaping a DataFrame of your own (interactive)
col = input('Type one of these to try melt or stack: month, year, passengers: ')
if col.strip().lower() == 'month':
    melted = df.melt(id_vars='month')
    print(melted.head())
elif col.strip().lower() == 'year':
    melted = df.melt(id_vars='year')
    print(melted.head())
elif col.strip().lower() == 'passengers':
    stacked = df.set_index(['month','year']).stack()
    print(stacked.head())
else:
    print('Try again with month, year, or passengers.')