notebooks 4, 5 etc.
Closed this issue · 2 comments
These notebooks would benefit from just loading stuff that has already been done (such as clustering, dimension reduction etc) and visualizing it. That way the notebooks run faster, you can reduce a lot of redundant code and the user gets familiar with the load/save functionality.
Maybe we can ship the notebooks with a few saved standard results for these 3 systems? When loading them one should point to the notebook that computes these results in the first place.
Maybe we can ship the notebooks with a few saved standard results for these 3 systems? When loading them one should point to the notebook that computes these results in the first place.
Providing precomputed results is a good idea and quite easy to do.
I do wonder, though, how much we should make the user repeat common tasks and go through the usual validation steps.