Calculating the missing values in AHP matrix
I have an incomplete AHP matrix, and I want to define the missing data using JMP, which method of screening values do you advice me to use, normal imputation, SVD imputation, or RPCA imputation. My second question as one part of the AHP matrix is equal to 1 divided on the second part. How can I tell JMP that to screen only the half of missed value which it located upper or under the diagonal axis and then to calculate the second half from the obtained result, or is there any symbol that I can use so JMP will not consider one part as a missing value but also will not affect the calculation of the other missing values
