I would like to perform stepwise regression on vglm objects. Since they are S4, method like `step()`

wouldn't work. I have found `step4()`

and `step4vglm()`

in `VGAM`

, which look like the equivalents of `step()`

, but for `S4`

objects. However, I kept getting the same error `Error in Deviance.categorical.data.vgam(mu = mu, y = y, w = w, residuals = residuals, : arguments 'y' and 'mu' must have at least 2 columns`

. I have trouble understanding and fixing this issue. Does anyone know what is wrong? Or is there a way to convert `S4`

to `S3`

and just use `step()`

to perform stepwise regression? I have attached a reprex if it helps.

```
##libraries
library(VGAM)
#> Loading required package: stats4
#> Loading required package: splines
# import data
dff<-list()
dff[[1]] <- data.table::data.table(
sbq_1 = c(1L,2L,1L,2L,2L,
1L,2L,2L,4L,2L,1L,4L,2L,3L,1L,4L,2L,1L,
3L,1L,4L,1L,3L,4L,2L,1L,2L,2L,2L,2L,
1L,4L,3L,4L,1L,2L,3L,4L,1L,1L,3L,3L,1L,
1L,2L,1L,1L,1L,1L),
age = c(53L,32L,47L,58L,
34L,29L,57L,25L,38L,29L,51L,50L,37L,33L,
41L,36L,49L,38L,32L,51L,37L,36L,37L,39L,
55L,55L,45L,25L,24L,27L,26L,57L,47L,30L,
37L,39L,38L,34L,38L,56L,39L,43L,49L,39L,48L,
44L,63L,61L,46L),
gender = c(1L,1L,2L,1L,1L,
1L,1L,2L,2L,1L,1L,1L,1L,1L,1L,2L,1L,1L,
2L,1L,1L,1L,1L,1L,2L,1L,1L,2L,1L,1L,
1L,1L,2L,1L,2L,1L,1L,2L,1L,1L,1L,2L,1L,
1L,2L,1L,1L,2L,1L),
ethnicity = c(2L,2L,2L,2L,2L,
2L,2L,2L,2L,2L,2L,1L,1L,2L,2L,2L,2L,2L,
2L,2L,2L,2L,2L,2L,2L,2L,2L,2L,2L,2L,
2L,2L,2L,2L,2L,2L,2L,2L,2L,2L,2L,2L,2L,
2L,2L,2L,2L,2L,2L)
)
ppom.1 <- vglm(sbq_1 ~ ., family=cumulative(parallel=F~ age), maxit=50, data=dff[[1]])
ppom.2 <- step4(ppom.1)
#> Start: AIC=135.61
#> sbq_1 ~ age + gender + ethnicity
#> Error in Deviance.categorical.data.vgam(mu = mu, y = y, w = w, residuals = residuals, : arguments 'y' and 'mu' must have at least 2 columns
summary(ppom.2)
#> Error in h(simpleError(msg, call)): error in evaluating the argument 'object' in selecting a method for function 'summary': object 'ppom.2' not found
```

^{Created on 2021-10-26 by the reprex package (v2.0.1)}