When resampling cluster size and cluster size was a covariate in one or more the ddf models supplied to mads sometimes crashed when trying to calculate abundance / density in dht.
On further investigation I discovered that the resample led to fitted values of zero in the ddf object.
Ddf was happy to fit the model and give a Pa of 0 and therefore an N of Inf but dht could not proceed
When resampling cluster size and cluster size was a covariate in one or more the ddf models supplied to mads sometimes crashed when trying to calculate abundance / density in dht.
On further investigation I discovered that the resample led to fitted values of zero in the ddf object. Ddf was happy to fit the model and give a Pa of 0 and therefore an N of Inf but dht could not proceed