01
GroupBy
⏱ ২০ মিনিট
df.groupby("city")["sales"].mean()
df.groupby("city").agg({"sales": "sum", "qty": "count"})
02
Merge & join
⏱ ২০ মিনিট
merged = pd.merge(orders, customers, on="customer_id")
combined = pd.concat([df1, df2], ignore_index=True)
03
Cleaning quiz
⏱ ২০ মিনিট
merge vs concat?
merge = column join; concat = stack rows/columns
04
GroupBy
⏱ ১০ মিনিট
df.groupby("city")["sales"].agg(["sum", "mean", "count"])
df.groupby(["region", "product"]).size()05
Merge & join
⏱ ১০ মিনিট
pd.merge(orders, customers, on="customer_id", how="left")
pd.concat([df2024, df2025], ignore_index=True)06
GroupBy quiz
⏱ ১০ মিনিট
left vs inner join?
left = keep all left rows; inner = only matching keys
groupby().apply() vs agg()?
agg named functions; apply custom per-group logic
duplicate keys merge-এ?
Cartesian explosion — dedupe বা validate keys আগে