The expected date of release of the SAXSGUI data analysis software is Monday the 5th of November.

… the software release schedule, that is.
Sorry about this. I cannot provide any software at the moment because more work on debugging the program needed to be done at the last minute.

Providing people with buggy software is a sure-fire way to get them to chuck it in the trash right away. The Google way is always available, we could call it a Beta until eternity. However, the Google way has also upped the standards for Beta software, so we’d have to have a severely debugged program at the very least.

It looks like the 2D fitting function is both a little to slow for convenience and user experience, and a little too unstable. My recommendation would be to keep away from that for the time being, so probably it will only be available in the release for advanced users.

I would, however, like to have some input on the design of the meta-gui, the database software I was working on that could be used as a launcher for the SAXSGUI. Since it uses parts of the code written by Karsten Joensen, I need to discuss the release licensing terms first, and can therefore not provide you with a download-able copy right here, right now. I will, however, as soon as I can.
But perhaps you could give me some input on which fields should be present in a database sample display. So far I have the following fields:
- a centre calibration file (saxsgui)
- a q calibration file (saxsgui)
- a maskfile (saxsgui)
- a multi-line note field
- a primary beam intensity field (abs. intensity)
- a transmission field (abs. intensity)
- an electron density contrast field (abs. intensity)
- a volume field (abs. intensity)
- an autoprocess script file field (sasxgui)
It furthermore shows the resolution, filename and measurement time for each sample, for helping the recognition process.

My question to you is, what fields would you like to see in a database GUI? What information would make you recognise a sample faster? Would you like to see other things in this GUI?

Any input can be sent to me using the comments box below. With your help, this interface can become more useful for everyone. Thanks in advance.

After a meeting with the Man behind the Main code, we figured that we should have a release version containing much of the saxsgui code (perhaps without the 2D fitting functionality), out by next week.

So be prepared, here comes the Total Analysis System for measurements!

B.

Over the last week-end, I have delved into the wonders of the Graphical User Interface (GUI in short). As an exercise, and a means to vent some of my ideas, I have programmed a sample database in Matlab.

The sample database originally started as a sample chooser for the “saxsgui”, an analysis program mainly written by Karsten Joensen, of whom I spoke before. As such, it still functions, but in its functionality, it has enough flexibility to be able to serve as a sample database.

It contains preview images of the scattering patterns, a user editable note field, and other fields that help manage data for use by the “saxsgui”. Keeping in line with the user interface guidelines of simple programs, I have eliminated the need for a “save” button. Whenever something is edited, a save is made.

There are still a few issues to figure out though, amongst others, the interface to the “saxsgui” is not completely functional, the preview images are a little on the dark side, backwards compatibility is untested, and there might be some sample naming issues (i.e. it only accepts unique filenames of the scattering patterns as input).

But all in all, I’m pretty content about it, and I hope it will be included in the next “saxsgui” release. You will be notified of this release as soon as it appears on the web.

As this is an open source program, any help is greatly appreciated. Why not contribute to the database or the saxsgui?

In the past, it has been customary to use a method of “peeling off” information from the scattering patterns, as opposed to an attempt at fitting an entire scattering pattern at once [1,2]. This was customary, partly perhaps because of its methodic clarity, but perhaps also due to the lack of computing power. The latter issue has been lifted off our shoulders by the ever increasing availability of power.

Whilst the complete model fit method is indeed less robust when compared to the sequential analysis method, the complete model fit eliminates potential weighting of the fitting parameters. In other words, if dependent parameters are determined sequentially, the chosen sequence will influence the fitting results (do note that the method is perfectly acceptable for independent parameters).

Therefore, a complete model fit to the entire scattering pattern would be the preferred method as opposed to a sequential determination. Complex samples, however, require either a very good initial guess, or initial fitting parameters from a sequential determination.

As it is, we are dealing with many parameters in these ordered systems, all of which influence the scattering pattern significantly. Personally, I have found Matlab a little lacking in its solving skills in this respect, so the hunt is on for me to find a suitable fitting method.

Additionally, the results that I have seen published so far, are quite light on the statistical details. I myself am quite unfamiliar with the topic myself, but it would be nice to see some statistical information on the model fits with these results. Therefore I hereby personally pledge to include these in any relevant publications.

I am working with dr. K.D. Joensen on SAXS analysis software, well documented and written in Matlab, chosen for its comprehensibility. This will be available shortly, and allows a complete model fit. Statistical data extraction on the fit is mine still to implement.

References:
[1] N. Stribeck. Analysis of saxs fiber patterns by means of projections. In B. S. Hsiao, editor, ACS symposium series, volume 739 of ACS symposium series, pages 41–56. American Chemical Society, 2000.
[2] H. F. Zhang, B. Q. Yang, and Z. S. Mo. Small angle x-ray scattering study of structural parameters of the aggregated state in polyamide 110 + bmi. European Polymer Journal, 32:57–59, 1996.