Error: Unable to compute predicted values with this model. This error can arise when insight::get_data() is unable to extract the dataset from the model object

[zhaoliang0302]

Hi, I encountered the following error while attempting to calculate p-values for the interaction between two groups in a linear mixed-effect model using hypothesis_test.

The relevant code is provided below:

tmp
# A tibble: 2,702 x 7
       Y  YEAR   age duration_yrs SEX    PATNO Group
   <dbl> <dbl> <dbl>        <dbl> <fct>  <dbl> <fct>
 1    29     0  65.1        0.882 Male    3001 low  
 2    30     1  65.1        0.882 Male    3001 low  
 3    30     2  65.1        0.882 Male    3001 low  
 4    29     3  65.1        0.882 Male    3001 low  
 5    29     4  65.1        0.882 Male    3001 low  
 6    29     0  67.6        1.07  Female  3002 low  
 7    29     1  67.6        1.07  Female  3002 low  
 8    29     2  67.6        1.07  Female  3002 low  
 9    27     3  67.6        1.07  Female  3002 low  
10    24     4  67.6        1.07  Female  3002 low  
# i 2,692 more rows
# i Use `print(n = ...)` to see more rows

LME_model <- lmer(Y~YEAR*Group+age+duration_yrs+SEX+(1+YEAR|PATNO), tmp)
mydf <- ggpredict(LME_model, c("YEAR", "Group"))
plot(mydf)

hypothesis_test(mydf)
# Error: Unable to compute predicted values with this model. 
# This error can arise when `insight::get_data()` is unable to extract the dataset from the model 
# object, or when the data frame was modified since fitting the model. You can try to supply a 
# different dataset to the `newdata` argument. If this does not work, you can file a report on the 
# Github Issue Tracker: https://github.com/vincentarelbundock/marginaleffects/issues.

[strengejacke]

Do you have a reproducible example? Furthermore, have you tried to convert the tibble tmp into a real data frame (using as.data.frame())?

[zhaoliang0302]

Thanks for your rapid reply! tmp data file is uploaded: https://www.dropbox.com/scl/fi/brl335xg23w1e3u6v224c/tmp.rda?rlkey=xyutctt6q5yd4yl8qoy42ihlb&dl=0

[strengejacke]

For some reason, dropbox server don't seem to work (temporarily?). Could you zip the file and add as attachment to this thread?

[zhaoliang0302]

tmp.zip
Thanks a lot!

[strengejacke]

hm, can't reproduce, works for me:

library(easystats)
#> # Attaching packages: easystats 0.6.0.13
#> ✔ bayestestR  0.13.1.7     ✔ correlation 0.8.4.9000
#> ✔ datawizard  0.9.0.1      ✔ effectsize  0.8.6.1   
#> ✔ insight     0.19.5.13    ✔ modelbased  0.8.6.4   
#> ✔ performance 0.10.6       ✔ parameters  0.21.2.6  
#> ✔ report      0.5.7.11     ✔ see         0.8.0.5
library(ggeffects)
#> 
#> Attaching package: 'ggeffects'
#> The following object is masked from 'package:easystats':
#> 
#>     install_latest
library(lme4)
#> Loading required package: Matrix

load("c:/Users/mail/Desktop/tmp.rda")

LME_model <- lmer(Y ~ YEAR * Group + age + duration_yrs + SEX + (1 + YEAR | PATNO), tmp)
mydf <- ggpredict(LME_model, c("YEAR", "Group"))
plot(mydf)

hypothesis_test(mydf)
#> # Linear trend for YEAR
#> 
#> Group    | Contrast |        95% CI |     p
#> -------------------------------------------
#> low-high |    -0.10 | [-0.24, 0.04] | 0.150

^{Created on 2023-10-12 with reprex v2.0.2}

Can you try following two steps:

1. easystats::install_latest(force = TRUE)
2. ggeffects::install_latest(force = TRUE)

and then retry?

[zhaoliang0302]

It works after install_latest(force = TRUE). Thanks for your time!