I've had this error when running somewhat complex trees on larger datasets. Some combination of these three actions usually allows the tree to visualize:
- reduce the number of predictors
- sample to reduce the number of input records into the tree algorithm module
- filtering to reduce the dataset complexity before input into the tree algorithm module (filter out all columns that are not dependent/independent columns in tree)
I'd start with (3) and see if that helps, since it doesn't affect the analysis, and then (2).