11. Reshape and Wide/Long Data
11.4 stack / unstack (Brief)
s = pt.stack() # Series with MultiIndex
back = s.unstack() # toward wide again
Useful when you already have a MultiIndex from groupby.
Practice task
Melt wide, then use pivot_table to get month as rows and city as columns (transpose the usual view). Which city-month is highest GMV?
Thodkyaat sangaycha tar (quick recap)
- Long form for analysis; wide for human grids
- melt ↔ pivot_table are the daily reshape pair
- Prefer pivot_table when duplicates need aggregating
Samajla ka? Aata pudhe jaauya Exploratory Data Analysis (EDA).