Unexpected results when using np.clip with masked arrays in NumPy 1.24.0
I'm working through a tutorial and Hey everyone, I'm running into an issue that's driving me crazy. I'm experiencing unexpected behavior when using `np.clip` on a masked array in NumPy 1.24.0. I have a masked array that contains both valid and invalid (masked) entries, and I expected `np.clip` to only apply the clipping operation to the unmasked entries. However, it seems to be affecting the masked values as well, leading to results that are not as anticipated. Here's a minimal example of what I've tried: ```python import numpy as np # Create a masked array with some values masked data = np.ma.array([1, 2, 3, 4, 5], mask=[0, 0, 1, 0, 1]) # Attempt to clip the values between 2 and 4 clipped_data = np.clip(data, 2, 4) print(clipped_data) ``` I expected the output to only show the valid values clipped between 2 and 4 while maintaining the mask, but what I actually got was: ``` [2 2 -- 4 --] ``` This suggests that the masked values are being set to 2 and 4, which is not what I intended. I've also tried using `data.compressed()` to see if that changes the outcome: ```python compressed_data = data.compressed() clipped_compressed = np.clip(compressed_data, 2, 4) print(clipped_compressed) ``` This gave me the correct results for the unmasked values, but I lose the masked structure which I need for further processing. Is there a way to achieve the desired outcome using `np.clip` while keeping the masked values intact? Any insights or best practices for handling masked arrays in this context would be appreciated. Am I missing something obvious? This issue appeared after updating to Python 3.10. I'm open to any suggestions.