Instead of asserting that the data to create 2D diagrams is a tensor, and thus enforcing the same number of observations per diagram, we can allow a list of tensors instead. As long as the number of treatments is identical, I do not see a necessity to enforce a tensor. This should be backward-compatible since len also works on tensors.
Instead of asserting that the data to create 2D diagrams is a tensor, and thus enforcing the same number of observations per diagram, we can allow a list of tensors instead. As long as the number of treatments is identical, I do not see a necessity to enforce a tensor. This should be backward-compatible since
len
also works on tensors.