Detect Label Leakage via Temporal Check
~12 mincode completion
Implement find_leaking_indices(feature_times, label_times) returning a sorted integer array.
Examples
One leak at index 1 where feature_time > label_time
- Input
- find_leaking_indices([1, 5, 3], [3, 4, 7])
- Output
- [1]
All features recorded before labels: no leakage
- Input
- find_leaking_indices([1, 2, 3], [4, 5, 6])
- Output
- []
Feature time exactly equal to label time counts as leakage
- Input
- find_leaking_indices([3, 3, 3], [3, 4, 2])
- Output
- [0, 2]
Hints
Hint 1
picks between two values elementwise without branching.
Hint 2
Watch for this: used strict greater than missing equal case.
Requirements
feature_times: 1D array of feature observation timestampslabel_times: 1D array of label observation timestampsReturn Sorted 1D integer array of leaking sample indices.
Constraints
Allowed library: NumPy only
Time limit: 200 ms, Memory: 64 MB
Where this shows up
~12 min
••••••••••••••••
8 employers weight this skill
3 big tech firms, 2 enterprise vendors, 1 defense company, 1 AI product company, 1 health and bio company. Top match scores 87.
Python
import numpy as np
def find_leaking_indices(feature_times: np.ndarray, label_times: np.ndarray) -> np.ndarray:
"""
Return sorted indices where feature_times[i] >= label_times[i] (temporal leakage).
Args:
feature_times: 1D array of feature observation timestamps
label_times: 1D array of label observation timestamps
Returns:
Sorted 1D integer array of leaking sample indices.
"""
# YOUR CODE HERE
pass