Data Preprocessing
Beginner
Clip Features to a Range
Beginner
~8 min
code completion
Clipping Features
Real-world data often contains outliers — extreme values that can destabilize model training. One simple defense is clipping: capping values at a minimum and maximum threshold.
NumPy provides this as a single call:
X_clipped = np.clip(X, low, high)
Your task:
Implement clip_features(X, low, high) that clips all values in X to the range [low, high].
Example Tests
Clip negatives and values above 10
Input: {"X":[-5,0,3,7,12],"low":0,"high":10}
Expected: [0,0,3,7,10]
Float bounds
Input: {"X":[1.5,2.5,3.5],"low":2,"high":3}
Expected: [2,2.5,3]
Clip large absolute values
Input: {"X":[100,50,0,-50,-100],"low":-30,"high":30}
Expected: [30,30,0,-30,-30]