CODEWITHSUNDEEP
rangeyaxissjplot
Adrianna Elihu
Adrianna Elihu

Reputation: 11

sjPlot- plot_model: how do I adjust my y axis?

I'm plotting 3 continuous variables (pland_herb, PD_woody, PD_herb) with my response variable which is Richness. I want to show the 3-way interaction that is occurring over 3 panels, same as this example below:

3-way interaction graph from regression model

However, I'm using a different code to get it:

plot_model(mod81, type = "int", mdrt.values = "all", terms = c("PD_Woody","PD_Herbaceous", "Pland_Herbaceous"))

Here is the resulting graph:

3-way interaction graph-messed up

How do I change the y-axis for the range of numbers to be much smaller? They're only from 0-45, not in the 1000's.

I also tried changing each of the 'type' options and that didn't fix anything.

I found another person's answer by Allen Cameron, which is almost helpful, however I don't want percentage, just whole numbers: Editing the y axis on a sjplot in R (plot_model) How can I alter that code?

Any help is appreciated! Thank you

Upvotes: 0

Views: 157

Answers (1)

Adrianna Elihu
Adrianna Elihu

Reputation: 11

I've tinkered with ggeffects code instead and got a plot closer to what I want! Hooray!

Here is the code:

dat_categorical <- predict_response(mod81, terms = c("PD_Woody","PD_Herbaceous [-1, 0, 1]", "Pland_Herbaceous [-1, 0, 1]"))
plot(dat_categorical, breaks = seq(0, 22, 2), limits = c(0, 22), ci_style = "ribbon")

corrected 3-way interaction plot!

Now my question is how do I fix the confidence interval ribbons, that seem to be stopping in the left and right panel?

Thanks again

Upvotes: 0

Related Questions