Hi , as far as I understand the inverse of multivariate distributions are not uniquely defined. If you would like to determine the cumulative inverse of a slice of data, I would recommend taking the data along the line of interest and either using:
1. The "norminv" function if the mean and standard deviation are known
2. Use the "fitdist" and "icdf" in combination if you would have to first determine what the mean and standard deviations are to begin with.
