25. How to Rank and Running Totals in pandas (rank, cumsum, window)

25.2 rank: Place Among Peers

By Ravindra Bagale, trainer · LinkedIn · Last updated: 24 October 2026

Why. Two Pune orders both ₹210. Is that rank 1 and 1, or 1 and 2? Interviewers love this. Managers hate silent ties.

What. Series.rank(ascending=False, method=...)

How.

Worked example A — Dense rank on baskets.

import pandas as pd

df = pd.DataFrame({
    "order_id": ["BLK-1", "BLK-2", "BLK-3", "AMN-4", "BLK-5"],
    "city": ["Pune", "Pune", "Nashik", "Kolhapur", "Pune"],
    "amount_inr": [210, 210, 95, 180, 150],
    "customer": ["Ruhi Bagale", "Amir", "Zoya", "Ravina", "Salman"],
})
df["rank_dense"] = df["amount_inr"].rank(ascending=False, method="dense")
df["rank_min"] = df["amount_inr"].rank(ascending=False, method="min")
df["rank_avg"] = df["amount_inr"].rank(ascending=False, method="average")
print(df.sort_values("rank_dense")[
    ["order_id", "amount_inr", "rank_dense", "rank_min", "rank_avg"]
])

What you should see. Both ₹210 rows share dense rank 1; next amount gets 2. min leaves a gap (next is 3). average shows 1.5.

Worked example B — Rank low delivery minutes (faster is better).

df = pd.DataFrame({
    "order_id": ["BLK-10", "BLK-11", "AMN-12"],
    "city": ["Solapur", "Nagpur", "Pune"],
    "delivery_mins": [11, 18, 11],
})
df["fast_rank"] = df["delivery_mins"].rank(ascending=True, method="dense")
print(df.sort_values("fast_rank"))

Worked example C — Interview one-liner.

“I use rank(ascending=False, method='dense') on GMV so ties share a place and the next city is still #2.”