How to transform least square means from Turkey table after box cox transformation

[#1]

Hello Everyone,

I am running a three-way ANOVA with data expressed in percentage, but the data set is not normal. Since I have zeros, I first add a constant 0.00001, then I fit the model, and in the least square fit window, I used the box cox transformation to do a better fit. I know if I ask to save the predicted values, it will save the back-transformed data, but is there an option for JMP to save the back-transformed data for the least square means and confidence interval generated for multiple comparisons?