[R-sig-ME] Multiple comparisons and post hoc tests on glmmTMB model, with the multcomp package.
Julian Gaviria Lopez
Ju||@n@G@v|r|@Lopez @end|ng |rom un|ge@ch
Mon Aug 26 18:43:31 CEST 2019
Hi,
I have the next model:
zipoisson <- glmmTMB(Observations ~ CAP * Condition + (1|ID), data=mDATA, ziformula=~ CAP * Condition , family=poisson)
Which result is:
Family: nbinom1 ( log )
Formula: Observations ~ CAP * Condition + (1 | ID)
Zero inflation: ~CAP * Condition
Data: mDATA
AIC BIC logLik deviance df.resid
4798.4 4979.5 -2365.2 4730.4 1486
Random effects:
Conditional model:
Groups Name Variance Std.Dev.
ID (Intercept) 0.003643 0.06036
Number of obs: 1520, groups: ID, 19
Overdispersion parameter for nbinom1 family (): 0.681
Conditional model:
Estimate Std. Error z value Pr(>|z|)
(Intercept) 1.09079 0.11695 9.327 < 2e-16 ***
CAPinsC5 -1.03221 0.32151 -3.210 0.00133 **
CAPpreC1 -0.10200 0.17408 -0.586 0.55794
CAPpreC5 -0.50219 0.21082 -2.382 0.01722 *
Conditionaff -0.17668 0.17403 -1.015 0.31000
Conditionneu -0.12073 0.18997 -0.636 0.52509
Conditionpneu -0.25180 0.19119 -1.317 0.18784
CAPinsC5:Conditionaff 1.17944 0.36350 3.245 0.00118 **
CAPpreC1:Conditionaff 0.40988 0.23992 1.708 0.08756 .
CAPpreC5:Conditionaff 0.33029 0.29746 1.110 0.26685
CAPinsC5:Conditionneu 0.75961 0.39673 1.915 0.05553 .
CAPpreC1:Conditionneu -0.05464 0.27420 -0.199 0.84205
CAPpreC5:Conditionneu 0.57299 0.28324 2.023 0.04308 *
CAPinsC5:Conditionpneu 1.01513 0.42694 2.378 0.01742 *
CAPpreC1:Conditionpneu 0.14104 0.27380 0.515 0.60647
CAPpreC5:Conditionpneu 0.22652 0.33243 0.681 0.49562
---
I want to apply the Multiple comparisons and post hoc tests proposed in:
https://cran.r-project.org/web/packages/glmmTMB/vignettes/model_evaluation.html#mumin
but it is not clear to me what stands for "components" and "period"
Thanks in advance for any input.
Julian Gaviria
Neurology and Imaging of cognition lab (Labnic)
University of Geneva. Campus Biotech.
9 Chemin des Mines, 1202 Geneva, CH
Tel: +41 22 379 0380
Email: Julian.GaviriaLopez using unige.ch
[[alternative HTML version deleted]]
More information about the R-sig-mixed-models
mailing list