I have been using the BCES fitting method of Akritas & Bershady (1996) to perform linear regression on data that has uncertainties on both X and Y (with bootstrapping). Unfortunately, the code made available by Akritas et al. has a few bugs which prevented me from getting it compiled with gfortran on my Mac.
I corrected those bugs in the fortran code and now I am able to compile and run the BCES routine without problems.
If you are interested in the patched BCES fortran routine, let me know and I will be glad to share it with you.
