Hi everyone,
I have values of prevalence of different species of canines for an infectious agent. I would like to predict what is the prevalence estimated for a raster of same dimensions of an environmental raster. The problem is when a unique individual was sampled (study case or hit by traffic) and this animal was infected the prevalence is maximum. I'm thinking to model the estimation of prevalence weighting how many individuals were sampled.
This is a bad idea use prevalence as response variable or is better to model a counting of individuals infected?
Is a prior contact with Bayesian models, therefore, I thanks for whatever assistance you might provide.
Bayesian is a study of sets . this means prevalence as an indicator or measure for total number of cases which includes some cases in the intersection set. since bayesian looks at predictors of the intersection based on the the sets participating to make it. we can change the subject of the formula so that prevalence( a measure of a set) is an outcome, if the other parameters are available.
Thanks Douglas Singini , Thanks Douglas Singini , I have a column for number of animals sampled and other for number of animals infected. How do you suggest to calculate the subject (prevalence)?
modeling is not my thing but can assess fitness of a model. if you have you data put it into a data set and use a programme like stata to apply regression and create so called DO files. You tube has some good tutorials for beginners. beauty about stata is that its more versatile as you will have to write the commands yourself.
- Badges
- Science topic
- Similar topics
- Medicine
- Epidemiology