Closed ercius closed 5 years ago
There was a recent change to this. It fixed some issues, but now many data sets will not load. Often I get an error that the step size can not be zero. This might happen for data smaller than 1e6. Also, I dont think this will work for a single image which is not 3D. Why not just use numpy slicing such as data[::10,::10]? That seems much simpler.
This is fixed. Its very fast even for fairly large datasets.
The new version of DynImageView.quickMinMax() made the contrast estimation significantly slower for large datasets. The entire data set is loaded into memory while searching for the min/max which takes a long time for datasets in the GB range.