Fix the wilsonepsilon thanks to Anders maths skills!
implementing ml in terms of "functors"
Tidied text to pass through latex without serious errors
Added .chi file reading
Added single crystal data type DILS objects (no overlap)
Checking for zero diagonal elements, translation and rotation functions and choosing particular sets of intensities
e^2 on Wilson epsilon included as in paper for Z.Krist
Testsuite starts
Initial revision
First step in translation functions - computing structure factors from model in P1 and applying symmetry. Next is to apply translations and carry out a translation search...
imagedata types for 2d peak fitting and starting on rotation fn
Summodels, product models...
Gui still a mess, but now sidetracked by translation functions
Gaussian approximation of likelihood implemented. This should be the version which matches the computations done for EPDIC 10 (2006).
Translation search seems to work(!)
I think the 2D weight matrix is now used correctly in the optimiser
Continuing with gui, was trying to work out how to instantiate things on the fly.
Re-factored single peak fitting code to separate model from optimiser
Work started on gui
Included Anders likelihood stuff - theory there, code next...
fix for g77