When running sparta across most taxonomic groups, typically half the species fail to meet the data thresholds. This can waste a lot of time. I recommend adding an if statement around line 216 of occDetFunc so that the majority of the function is skipped if the data fail to meet a certain data threshold. Ideally these thresholds could be supplied as a thresholds argument to occDetFunc. However, coding for the decision tree is more complicated, so we may instead have to give three options: an integer to specify the number of records (e.g. 50, as per Charlie's work), then ROT_equalWeight and ROT_highSpecificity as alternatives.
drnickisaac
commented
3 years ago
When running sparta across most taxonomic groups, typically half the species fail to meet the data thresholds. This can waste a lot of time. I recommend adding an
ifstatement around line 216 ofoccDetFuncso that the majority of the function is skipped if the data fail to meet a certain data threshold. Ideally these thresholds could be supplied as athresholdsargument tooccDetFunc. However, coding for the decision tree is more complicated, so we may instead have to give three options: an integer to specify the number of records (e.g. 50, as per Charlie's work), thenROT_equalWeightandROT_highSpecificityas alternatives.