Sometimes resampling is unnecessary (for example, in an ensemble custom predictor where all models are trained with the same spacing and there's a custom initial resampler). In such cases, it's useful to set no_resampling_data_or_seg_to_shape to avoid dealing with resampling altogether. Additionally, I noticed several unused imports and comments, which I removed to keep the code clean and well-organized.
Sometimes resampling is unnecessary (for example, in an ensemble custom predictor where all models are trained with the same spacing and there's a custom initial resampler). In such cases, it's useful to set
no_resampling_data_or_seg_to_shapeto avoid dealing with resampling altogether. Additionally, I noticed several unused imports and comments, which I removed to keep the code clean and well-organized.