zero inflated and repeat cells
------------------------------ Message: 2 Date: Tue, 10 Jul 2012 15:40:21 -0500 From: "Stephanie L. Simek" <ssimek at CFR.MsState.Edu> To: <r-sig-mixed-models at r-project.org> Subject: [R-sig-ME] zero inflated and repeat cells Message-ID: <736BA42933F6C84698DBA828209ED41E01AD1026 at mail.cfr.msstate.edu> Content-Type: text/plain Dear R users, I have count data that illustrates the presences or absence of individuals in my study population. I created a grid of cells across the study area and calculated a count value for each individual per season per year for each cell. The count value is the number of times an individual was present in each cell. My data is set up with the cell ID repeated for each individual per season per year. This format results in 71,000 records (of which 925 have a count value >0 and 70,075 have a count value = 0). I would like to run mixed effects model with individual and year as random effects but I have been advised two things:
I don't know whether the approach sketched above is good, bad or pragmatic (I would need to know more details, but I did raise my eyebrow), but as to your specific questions below: 1. Not sure if I would agree. It all depends on where the zeros are, and whether some of the covariates can be used to model the zeros....in which case even a Poisson GLM may do the job. But a ZIP is probably ok. 2. You need to think very careful where any dependency structures are in your experiment. Within a cell...between cells? Without further information I can't say. Don;t you have spatial correlation between these cells? Alain
1. The data are too extremely zero inflated therefore a zero inflated Poisson model will not work. 2. The model cannot be run because the grid cells are repeated for each individual per season per year. I am told the model doesn't recognize that Cell ID #1 for individual "A" is the same cell for Cell ID #1 for individual "B". Does anyone know if either or both of these points are true? Also, does anyone have any recommendations. Thank you, -Stephanie ------------------------------------------------------- Stephanie L. Simek Carnivore Ecology Lab Forest and Wildlife Research Center Mississippi State University Box 9690 Mississippi State, MS 39762 Cell: (850) 591-1430 Email: ssimek at cfr.msstate.edu [[alternative HTML version deleted]] ------------------------------
_______________________________________________ R-sig-mixed-models mailing list R-sig-mixed-models at r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models End of R-sig-mixed-models Digest, Vol 67, Issue 13 **************************************************
Dr. Alain F. Zuur First author of: 1. Analysing Ecological Data (2007). Zuur, AF, Ieno, EN and Smith, GM. Springer. 680 p. URL: www.springer.com/0-387-45967-7 2. Mixed effects models and extensions in ecology with R. (2009). Zuur, AF, Ieno, EN, Walker, N, Saveliev, AA, and Smith, GM. Springer. http://www.springer.com/life+sci/ecology/book/978-0-387-87457-9 3. A Beginner's Guide to R (2009). Zuur, AF, Ieno, EN, Meesters, EHWG. Springer http://www.springer.com/statistics/computational/book/978-0-387-93836-3 4. Zero Inflated Models and Generalized Linear Mixed Models with R. (2012) Zuur, Saveliev, Ieno. http://www.highstat.com/book4.htm Other books: http://www.highstat.com/books.htm Statistical consultancy, courses, data analysis and software Highland Statistics Ltd. 6 Laverock road UK - AB41 6FN Newburgh Tel: 0044 1358 788177 Email: highstat at highstat.com URL: www.highstat.com URL: www.brodgar.com