explainPredictions always uses maximum number of trees

[ChoGit]

Hi David and thank you for this useful explainer. I must point out that I stepped into an issue with explainPredictions function.
The number of iterations of for loop is defined by lines
54: nodes = predict(xgb.model,data,predleaf =TRUE)
and
61: num_trees = ncol(nodes)
None of this lines takes into account the number of trees used to built the explainer variable and so num_trees is always equal to the total number of trees present into the model, even thou when I called buildExplainer I set the input variable n_first_tree to a lower value. In this case, after n_first_tree iterations the for loop start adding NAs to preds_breakdown producing a NAs-Only output.