Feature Request: Change order of reference level in ggcoef_compare()

[jerrodanzalone]

I find your ggocef_compare() such a useful tool for comparing models. Is there a way to retain the factor ordering for the reference levels?

d <- as.data.frame(Titanic)

d$Sex <- factor(d$Sex, levels=c("Male", "Female"))
d$Sex=relevel(as.factor(d$Sex),ref="Male")

m1 <- glm(Survived ~ Sex + Age, family = binomial, data = d)
m2 <- glm(Survived ~ Sex + Age + Class, family = binomial, data = d)

models <- list("Model 1" = m1, "Model 2" = m2)

ggcoef_compare(models, exponentiate = TRUE)

Which produces:

This differs from ggcoef_model(), which retains the specified order:

d <- as.data.frame(Titanic)

d$Sex <- factor(d$Sex, levels=c("Male", "Female"))
d$Sex=relevel(as.factor(d$Sex),ref="Male")

m1 <- glm(Survived ~ Sex + Age, family = binomial, data = d)    

ggcoef_model(m1, exponentiate = TRUE)

There is probably a straightforward way to do this, but I'm struggling to find it. Additionally, if I try to remove the reference row, I receive an error:

d <- as.data.frame(Titanic)

d$Sex <- factor(d$Sex, levels=c("Male", "Female"))
d$Sex=relevel(as.factor(d$Sex),ref="Male")

m1 <- glm(Survived ~ Sex + Age, family = binomial, data = d)
m2 <- glm(Survived ~ Sex + Age + Class, family = binomial, data = d)

models <- list("Model 1" = m1, "Model 2" = m2)

ggcoef_compare(models, exponentiate = TRUE, add_reference_rows = FALSE)
```

Error in eval_tidy(dot, data = mask): object 'reference_row' not found

Thanks!

[larmarange]

It should be fixed by #24

[jerrodanzalone]

@larmarange Fantastic. Thank you!