I model the cumulative number of cases of Ebola using the Richard's model. I used both R and SAS and both gave the same estimates for parameters. However, I can not get the 95% confidence interval using R (growthmodels package), that is why I need to use NLMIXED in SAS to have 95% CI. If any one could show me how to get the 95% CI for the prediction using R. One more concern, how do I calculate the final epidemic size. Based on the interpretation of the parameters, it seems that the final epidemic size is equal to parameter alpha (I used the model specification in the paper of Fekedulegn D. et al., 1999 "Parameter Estimation of Nonlinear Growth Models in Forestry"). Thank you for any help.
Sorry but it seems that I don't see your answer. Is it an error?
Thank you.
Thank you. The Richards' model in this case underestimates the cases in June2015 (the data I got only had the cumulative cases until 18Mar2015). The final epidemic size estimate is also less than the cases recorded on 18Mar2015, that makes me a little bit worry. However, I tried other models (Gompertz, logistic, Negative Exponential, Chapman Richards and Mitcherlich), it seems that the Richard's model performs the best in terms of AIC.
- Badges
- Science topic