jyu-iba / jabs Goto Github PK

Documentation is needed

• Built-in interactive help, especially usage hints should be checked, expanded and improved
• The code needs better commenting (Doxygen) and documentation generated by these comments
• Physics models need to be explained. A manuscript is under preparation, but an additional continuously updated document is needed
• More examples, along with explanations, should be produced

When fitting e.g. the detector calibration, JaBS currently adds fit parameters to all detectors. This is a sane default, but there should be an option to add a fit parameter of one particular detector.

Spectra are simulated also outside ROIs when fitting. In particular computing time could be saved by:

• Calculating the minimum energy in a simulation based on lowest energy in a ROI
• Convolution of spectra by resolution is only required for channels inside ROIs
• In multidetector mode there can be entire spectra that have no ROIs at all, in which case simulation can be skipped entirely.

After the fit has converged all spectra should be simulated completely (as if running simulate)

Hi,
is it planned to add the brew installation for the code similar to jibal?
This can be really helpful for quick update to the new version...
Thanks!
Ilya

Transmission geometry is currently not supported, but it should be possible to implement. Escaping reaction products should not assume they are going towards the surface.

Concentrations (of sample_model) are always renormalized to 100% before simulation. This means that the fitted variable is not actually concentration, and the renormalization will change the value that was changed by fitting routine. This may confuse the GSL Levenberg-Marquardt algorithm (unknown).

Some new variable could be introduced for actual concentration fitting, instead of the actual concentration value or we should only normalize the sample (simplified isotopic concentration table).

In any case, the fit uncertainty of final value needs to be checked, since it is affected by the normalization.

Compatibility with

• IDF
• RUMP

Would be useful. IDF files could be turned into data and an autogenerated script file that sets the parameters described by the XML file. This would be best done by a program other than JaBS itself.

RUMP commands and files could be supported natively, but the question is should they be?

The graphical user interface is currently an editor. It could also include tools and wizards to generate scripts.

For example, the user could pick an element/isotope using a table of elements / combobox etc., set the energy, scattering angle etc. with a spin box and generate experimental details script, which could be then appended to the editor.