Wavelet transform adds columns to resulting array

[tsalo]

In the test dataset, there are 239 volumes. After apply the wavelet transform and then the inverse transform, there are 240, even when nothing's been done to the data.

Weirdly, np.allclose between the original data and the first 239 volumes of the transformed data passes.

It seems like the extra column is added when the number of columns in the raw data is >200. Fortunately, based on PyWavelets/pywt#564 (comment), it seems like that final column can just be trimmed out!