I have been evaluating the quality of modelfit with the easystats package (0.7.1.3). Since I have mixed models I use the r2_nakagawa() function and since the last update I did (easystats::install_latest(force = TRUE)) this function gives me strange values.
For example for this model:
<code>m2<-glmmTMB(FvFm ~ hora*temperatura + (1|Experimento), data=d, family=beta_family(link="logit"))
</code>
<code>m2<-glmmTMB(FvFm ~ hora*temperatura + (1|Experimento), data=d, family=beta_family(link="logit"))
</code>
m2<-glmmTMB(FvFm ~ hora*temperatura + (1|Experimento), data=d, family=beta_family(link="logit"))
I used to get these values:
<code>r2_nakagawa(m2)
# R2 for Mixed Models
Conditional R2: 1.000
Marginal R2: 0.973
</code>
<code>r2_nakagawa(m2)
# R2 for Mixed Models
Conditional R2: 1.000
Marginal R2: 0.973
</code>
r2_nakagawa(m2)
# R2 for Mixed Models
Conditional R2: 1.000
Marginal R2: 0.973
and now (with the latest update) it gives me these values:
<code># R2 for Mixed Models
Conditional R2: 1.341
Marginal R2: 1.306
</code>
<code># R2 for Mixed Models
Conditional R2: 1.341
Marginal R2: 1.306
</code>
# R2 for Mixed Models
Conditional R2: 1.341
Marginal R2: 1.306
I don’t know what’s wrong, but I have the latest version of easystats and performance and insight.
Someone is having these problems, I think it is not code.
Any help is welcome and appreciated!!
I tried to install previous versions of easystats, but the values for R2 did not change.