Error when plotting effects from model with multiple grouping variables

[SamDeCraemer]

Hi,

Let me know if what I am trying to do is simply not intended usage of ALASCA. Otherwise, here is my issue.

I was trying to get the effect plots of a models with a time component, two fixed effects to be plotted, and a random effect variable:

model <- ALASCA(ALASCA_tb,
formula = value ~ time*var1+ var2+ time:var2+ (1 | rndvar1),
                effects = c("var1+ time:var1", "var2+ time:var2"), 
                method="LMM",  
                participant_column="rndvar1",                 
                use_Rfast=F
)

When plotting the resulting model, I get this warning

Error in `[.data.table`(data_to_plot, , `:=`(group_data, do.call(paste,  : 
  Some items of .SDcols are not column names: [donation_type]

model$splot$model$effect_list$terms correctly lists the terms for each effect, but the function model_loaded$splot$plot_effect_score will incorrectly try to include the grouping term unique to the second effect for plotting the first effect. I traced it down to this line in the plot_effect_score function

data_to_plot[, `:=`(group_data, do.call(paste, c(.SD, 
        sep = "-"))), .SDcols = self$model$effect_terms[-1]]

The .SDCols names are gotten from model$splot$model$effect_terms which contains all terms of all effects, [-1] then assumes there is only one grouping term to be dropped, that is the first column. Perhaps instead self$model$effect_list$terms[[effect_i]][-1] could be used to make sure only the time term is kept?