Thomas de Boer

Hi all, I have been porting a fairly large piece of code from IDL to GDL, to run it on the computing cluster at my local Uni. This has mostly been straightforward and intuitive, but I am encountering a stubborn issue. The code tries to fit a full-sky Milky Way model to the observed night sky in a fine grid of Healpix, and does so by convolving the model with the observed uncerainties. The minimisation is done using an MCMC solver, so the model fitting function is called a lot. To do the convolving,...