scPred probabilities greater than 1

[Seandelao]

Hello,
Thanks for the great tool, it's been very helpful for my studies. I've noticed that when plotting the probabilities scores added to the meta data of my query seurat object, some of the cluster predictions will be greater than 1. It's like this for about half of my predicted clusters, while the others look great. Here are two examples of a normal looking probability calculation vs. one that is messed up. Any idea what may be happening?

[joseah]

Hi @Seandelao ,
this definitely shouldn't be happening. Do you have more information about the training models, error/warning messages?
Cheers,
José