Join Our 5-Week ML/AI Engineer Interview Bootcamp 🚀 led by ML Tech Leads at FAANGs
209. Mean aggregation
GeneralMean aggregation is a simple feature-engineering trick where you summarize a group of values with their average. In this task, you’ll build a function that creates per-group mean features and attaches them back to each row.
Requirements
Implement the function
Rules:
- Compute the mean for each group (g) as ( \mu_g = \frac{1}{|g|}\sum_{i \in g} x_i ).
- Append the corresponding group mean as a new last column to every row.
- Don’t use pandas; use only NumPy and built-in Python libraries.
- Keep the original row order unchanged.
- Output must be a NumPy array.
Example
Output:
| Argument | Type |
|---|---|
| rows | np.ndarray |
| value_idx | int |
| group_key_idx | int |
| Return Name | Type |
|---|---|
| value | np.ndarray |
Constraints
No pandas; only Python/NumPy allowed
Return NumPy array
Preserve original row order
Hint 1
Do two passes: first compute per-group totals and counts; second append the group mean to each row.
Hint 2
Use dictionaries keyed by rows[i][group_key_idx]: sums[key] += value, counts[key] += 1, then mean = sums[key]/counts[key].
Hint 3
When building output, don’t mutate input rows: new_row = list(row) then new_row.append(means[key]), preserving original order.
Roles
Companies
GeneralLevels
Tags
Python Editor