Advanced Data Transformations

Melt — Wide to Long Format

Transform columns into rows. Perfect for preparing data for plotting or modeling.

# Wide format: each category is a column
data_wide = {
    "city": ["NYC", "LA", "Chicago"],
    "2020": [100, 150, 200],
    "2021": [120, 160, 220],
    "2022": [140, 180, 250]
}

df_wide = pd.DataFrame(data_wide)
# Melt to long format
df_long = df_wide.melt(id_vars="city", var_name="year", value_name="sales")

Result: Each row represents one (city, year) pair.

Stack & Unstack

• Stack: Pivot rows → columns (opposite of melt)
• Unstack: Pivot index levels → columns

df.set_index(['region', 'product']).unstack()

MultiIndex (Hierarchical Index)

Index with multiple levels for organizing complex data:

df = pd.DataFrame({
    'value': range(6)
}, index=pd.MultiIndex.from_tuples([
    ('A', 'X'), ('A', 'Y'), ('B', 'X'), 
    ('B', 'Y'), ('C', 'X'), ('C', 'Y')
], names=['region', 'product']))

df.loc['A']        # Get all products in region A
df.loc['A', 'X']   # Get specific region-product

When to use: MultiIndex is powerful but can be confusing. Use it when you have hierarchical data (years > months > days).