You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Two small problems with the intPGOcc and spIntPGOcc functions in v0.3.0.
If a data set used in the integrated model only has site-level covariates in the detection model and does not have an even detection history (i.e., not all sites for that data set have the same number of replicate surveys), then the model will not fit properly. If you convert the site-level covariates to a full site x replicate matrix with the same missing data pattern as the detection-nondetection data, the model will work properly.
The summary function for a posterior predictive check performed on an integrated model will round the resulting Bayesian p-values to 0 or 1. To extract a Bayesian p-value from the resulting ppcOcc object (say its called object), you can run the following
# q iterates across the number of data sets in the integrated model
mean(object$fit.y.rep[[q]] > object$fit.y[[q]])
These problems are fixed on the development version, and will be updated on CRAN at the end of April 2022.
The text was updated successfully, but these errors were encountered:
Two small problems with the
intPGOcc
andspIntPGOcc
functions in v0.3.0.summary
function for a posterior predictive check performed on an integrated model will round the resulting Bayesian p-values to 0 or 1. To extract a Bayesian p-value from the resultingppcOcc
object (say its calledobject
), you can run the followingThese problems are fixed on the development version, and will be updated on CRAN at the end of April 2022.
The text was updated successfully, but these errors were encountered: