CCTBX

I am reaching the stage in xia2 where it would be really useful to be able to use crystallographic code from CCTBX. This will however incur another dependency. How do you feel about this?

Please cast your vote on the right!

XIA2 0.2.6.3 released

Many fixes made:

Changes since 0.2.6.2

Now works for images with long image numbers, as typically recorded on Rigaku X-Ray sets.

Signed off for operation with Rigaku Saturn and RAXIS IV detectors.

Now works with no input, e.g.
xia2 -project TG6623 -crystal X77788 -atom se /my/images/are/here

However this relies on your image headers being accurate and the images having some kind of recognisable format...

Repaired operation on ppc and intel macs - added required libraries to the installation and reset the library paths appropriately.

Slightly improved error reporting from xinfo file errors.

A great many other small changes so this is a worthwhile update!

Slightly delayed release

During testing of the next forthcoming release I have found a few cases where XIA2 works rather badly for nasty cases - e.g. gets stuck in a loop. The release 0.2.6.3 will come once I have fixed these, hopefully next week...

User interface

*finally* I have gotten around (with a great deal of help from Pete briggs) to putting together a CCP4i interface to xia2. This means no more messing around on the command line - you can just type everything into the GUI.

This will be included in CCP4 6.1 but may also be available before then...

xia2 for the *truly* lazy / busy

So, quite a few users have complained about the pain and suffering required to write the input file. Well, I have done away with it now then. You can still keep it - and I would advise this - but if you don't want to bother then you can try:

xia2 /path/to/images

and it will do it's best. If you are feeling slightly charitable you can give some clues:

xia2 -project SG901 -crystal X66252 -atom Se /path/to/images

If you have MAD data, it will pick this up. If you have multiple passes, it will pick this up. If you have SAD data you *have* to give it an atom else I don't know. You are allowed to lie.

If xia2 screws up just edit automatic.xinfo and you should be fine.

Usual commandline options like -debug -2d -3d -parallel etc still work.

XDS versions

The latest version of XDS (well, XSCALE really) seems to demand a slightlyt different format of XSCALE.INP file - I am working on this and will publish a patch here (and on the mailing list) as soon as I have one...

Error corrections

Currently working on a "proper" way to correct the chi-squared values (and hence the sd's) of intensities from Scala - first pass seems that it is much more effective and much quicker than the current run-scala-lots-remerging solution...