Page 18 in chapter 14. Step 3 reads 'take a sample from the residuals and invert these to obtain a set of pseudo-data'. What does it mean by inverting residual?
In this context inverting a residual means going back from the residual to a (pseudo-) data item. So if the residual was say the difference between the actual data and the expected data then you would add your "sampled" residual to the expected data value to get a new pseudo data value. Have a look at Step 2 on page 20 which shows you this in diagrammatic form.
Do you mean that "a" randomly selected residual is added to all the numbers in the actual triangle? and that this process is repeated to create many pseudo triangles?
The process is to sample from your residual distribution for each element of the upper left triangle. So not just a single residual for every element, but a (possibly) different residual for each element.
