Ravindra BagaleCourses & study guides मराठी Track your progress

36. K-Means Clustering in Python

36.4 India Customer / City Fictional Example

By Last updated:

Why. Wire a full tiny India example: fictional grocery customers across Maharashtra cities.

What. Fit KMeans(n_clusters=3), attach cluster ids, and summarise means per cluster.

How.

Copy-run customer groups

  1. Build numeric X.
  2. Fit KMeans with n_init set (avoids warning noise).
  3. Add cluster column.
  4. groupby mean to name the groups.
import pandas as pd
from sklearn.cluster import KMeans

df = pd.DataFrame({
 "customer": ["Ruhi", "Raja", "Rani", "Asha", "Vijay", "Meera", "Om", "Neha"],
 "city": ["Pune", "Kolhapur", "Nashik", "Pune", "Mumbai", "Nagpur", "Pune", "Nashik"],
 "orders_per_month": [2, 1, 9, 3, 11, 2, 10, 8],
 "avg_basket_inr": [240, 190, 910, 310, 1250, 220, 980, 870],
})
X = df[["orders_per_month", "avg_basket_inr"]]
km = KMeans(n_clusters=3, n_init=10, random_state=42)
df["cluster"] = km.fit_predict(X)
print(df[["customer", "city", "cluster"]])
print(df.groupby("cluster")[["orders_per_month", "avg_basket_inr"]].mean().round(1))

What you should see. Three buckets — e.g. light shoppers vs heavy baskets. City is shown for colour; it was not used as a feature here so distance stays numeric and simple.

Fictional data

Names and baskets are classroom fiction only.