Join Our 5-Week ML/AI Engineer Interview Bootcamp 🚀 led by ML Tech Leads at FAANGs
Back to Questions
88. Detecting duplicate features
easy
Generalsenior
Detect duplicate (or perfectly correlated) features during feature engineering so you can safely drop redundant columns. You’ll treat two features as duplicates if they are exactly equal for every row.
The duplicate check is defined as:
Requirements
Implement the function
python
Rules:
- Two features are duplicates only if their entire columns match exactly across all rows.
- Return all duplicate pairs
(feature_names[i], feature_names[j])fori < jthat are duplicates. - Use only NumPy and Python built-in libraries (no pandas).
- Keep the implementation in a single Python function.
- Key considerations: compare columns efficiently, keep output ordering stable, avoid converting to strings for comparison, handle multiple duplicates of the same feature, and return names (not indices).
Example
python
Output:
python
Input Signature
| Argument | Type |
|---|---|
| X | np.ndarray |
| feature_names | np.ndarray |
Output Signature
| Return Name | Type |
|---|---|
| value | np.ndarray |
Constraints
Use NumPy; no pandas
Return pairs in i<j scan order
Compare columns exactly; no string conversion
Hint 1
Slice columns directly from X as X[:, i].
Hint 2
Scan feature pairs in stable order with nested loops (i < j) and compare full columns.
Hint 3
Use np.array_equal(X[:, i], X[:, j]) for exact, row-wise identity; append [feature_names[i], feature_names[j]] immediately to preserve discovery order.
Roles
ML Engineer
AI Engineer
Companies
GeneralLevels
senior
entry
Tags
numpy
feature-engineering
duplicate-columns
array-equality
26 people are solving this problem
Python EditorLn 1, Col 1