Closes #992. Because cell clustering now takes a considerable amount of time, chance of intermediate failure also increases. There is no current way to recover intermediate cell training data if this happens, so this needs to be added.
How did you implement your changes
Add intermediate saving of the cell training data in the cell clustering notebook. This allows any data, along with any cluster labels, to be loaded in on restart.
Add overwrite functionality to the cell SOM assignment and cell consensus clustering functions. This will allow the user to redo these if they use different parameters.
What is the purpose of this PR?
Closes #992. Because cell clustering now takes a considerable amount of time, chance of intermediate failure also increases. There is no current way to recover intermediate cell training data if this happens, so this needs to be added.
How did you implement your changes
overwrite
functionality to the cell SOM assignment and cell consensus clustering functions. This will allow the user to redo these if they use different parameters.