0

I am currently doing mixed effects model analysis for a university project. I am currently trying to measure potential influences in my data using the following and become stuck with this error message in R.

alt.est.S <- influence(Smdl1,"subject")

Error in get(as.character(model@call)[3]) : object 'FALSE' not found.

My model is as follows:

Smdl1 <- lmer(midpoint~season+(1|subject),REML="FALSE")

Is there anything obvious here that stands out as reasoning for this error message? Any feedback would be really appreciated.

Julius Vainora
  • 47,421
  • 9
  • 90
  • 102
twstats
  • 1
  • I fit the model `mod <- lmer(Petal.Width ~ Petal.Length + (1|Species), iris, REML = FALSE)`, which has the same basic structure as yours does. the `influence()` function worked fine for me. I would assume that the error you're getting has something to do with your data, but I can't check unless you post the dataset. – Joseph Clark McIntyre Jan 06 '19 at 01:27
  • 1
    can you let us know `packageVersion("lme4")` please? – Ben Bolker Jan 06 '19 at 01:33
  • I had to upgrade `lme4` to get the `iris` example above to work. `install.packages("lme4")` – shuckle Jan 06 '19 at 02:31
  • the lme4 package version is 1.1.19 – twstats Jan 06 '19 at 02:37
  • I have just added "iris" into my model description and now produce a new error using influence() of: "Error in model@flist[[group]] : recursive indexing failed at level 2" I guess this would imply there is an issue with my dataset? – twstats Jan 06 '19 at 02:50

0 Answers0