Hi folks data structure: plots nested in sites, and observations nested in plots. both site and plots used as random factors When running a mixed model (lme4) I founded spatial autocorrelation on model residuals, thus, I tried nlme mixed models inclunding a spatial structure (e.g. corr=corExp(...)), since I have repeated measures at the same point (then, with the same coordinates) the package is unable to run, giving the error: Cannot have zero distances in "corSpatial" in nlme It seems to be a common and unsolved problem https://www.google.com/search?q=Cannot+have+zero+distances+in+%22corSpatial%22&ie=utf-8&oe=utf-8&client=firefox-b It is my first time here, sorry for any mistake in my post. Thanks vmaia
Cannot have zero distances in "corSpatial"
2 messages · Vinicius Maia, Thierry Onkelinx
Dear Maia, Add a correlation structure with nlme won't solve your problem. Is applies the correlation structure on the residuals **within** each finest group as defined by the random effects. The residuals **between** random effects group are assumed to be independent. The spatial structure is your design is probably defined by site and plot so this won't work. Have a look at the INLA or inlabru packages. They can model spatially correlated random effects. See https://www.muscardinus.be/2018/07/inlabru-bru/ for some examples. Best regards, ir. Thierry Onkelinx Statisticus / Statistician Vlaamse Overheid / Government of Flanders INSTITUUT VOOR NATUUR- EN BOSONDERZOEK / RESEARCH INSTITUTE FOR NATURE AND FOREST Team Biometrie & Kwaliteitszorg / Team Biometrics & Quality Assurance thierry.onkelinx at inbo.be Havenlaan 88 bus 73, 1000 Brussel www.inbo.be /////////////////////////////////////////////////////////////////////////////////////////// To call in the statistician after the experiment is done may be no more than asking him to perform a post-mortem examination: he may be able to say what the experiment died of. ~ Sir Ronald Aylmer Fisher The plural of anecdote is not data. ~ Roger Brinner The combination of some data and an aching desire for an answer does not ensure that a reasonable answer can be extracted from a given body of data. ~ John Tukey /////////////////////////////////////////////////////////////////////////////////////////// <https://www.inbo.be> Op do 6 dec. 2018 om 19:33 schreef Vinicius Maia < vinicius.a.maia at hotmail.com>:
Hi folks data structure: plots nested in sites, and observations nested in plots. both site and plots used as random factors When running a mixed model (lme4) I founded spatial autocorrelation on model residuals, thus, I tried nlme mixed models inclunding a spatial structure (e.g. corr=corExp(...)), since I have repeated measures at the same point (then, with the same coordinates) the package is unable to run, giving the error: Cannot have zero distances in "corSpatial" in nlme It seems to be a common and unsolved problem https://www.google.com/search?q=Cannot+have+zero+distances+in+%22corSpatial%22&ie=utf-8&oe=utf-8&client=firefox-b It is my first time here, sorry for any mistake in my post. Thanks vmaia [[alternative HTML version deleted]]
_______________________________________________ R-sig-mixed-models at r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models