Question on the estimated item probability to draw a profile plot

[Heeryung Choi]

Hi all, 

I am new to this MplusAutomation package, and it has been tremendously helpful. 

I am trying to follow Figure 1 and code listing 2 in the article "MplusAutomation: An R Package for Facilitating Large-Scale Latent Variable Analyses in Mplus" to draw profile plots.

If I understood correctly, the y axis of Figure 1 represents percentages of every value 1 versus value 0 in each of the categorical values. My question is what happens if I follow the code and draw plots using categorical data which has more than two categories. For example, in my data, some categorical variables have up to 4 categories. In this case, if I throw my data into the code, I get a profile plot that looks very similar to Figure 1, but I have no idea how to interpret this. Could anyone help me? If this plot/code is not the best way to compare different classes with categorical variables which are not binary, could anyone point me in the right direction? Thanks a lot!

Best,

Heeryung