The NEWS of the randomForest R library mention that version 4.5-13 fixed a bug in predict.randomForest() when newdata is a matrix with no rownames. Can anyone explain what were the consequences of this bug in the predictions? I think the bug was fixed with the new line if (is.null(rn)) rn <- keep in file predict.randomForest.R I got predictions with the new version using a RF trained before version 4.5-13 and I can't see differences... Thanks, jaslqa