Beam model should state probe type

[dagewa]

Spotted while processing some electron diffraction data where the detector material and thickness was specified:

$ dials.show integrated.refl | grep "qe"
| qe                         | 0.009                              | 0.009                              | 0.009                              |

The QE correction is assuming 200 keV photons in this case and thus multiplying up all the intensities by about 100. In truth, these are electrons and are all easily stopped in the detector. So, the QE correction at least needs to know the probe type so we can either switch it off or do something cleverer here.

The first step then is to make Beam models aware of X-rays, electrons and neutrons...

[graeme-winter]

https://manual.nexusformat.org/classes/base_classes/NXsource.html

The choice of the name probe is perhaps questionable for those of us of a certain age with a fondness for anarchic cartoons, but it is the de facto standard.

[graeme-winter]

For completenss, suggest we select the names from above as

neutron
x-ray
muon
electron
ultraviolet
visible light
positron
proton

I find it unlikely that we support positrons or muons in DIALS, but ...

[dagewa]

Actually the QE correction does not take in the Beam model. It should be possible to correct behaviour by setting an appropriate µ factor in the format class. So that's an orthogonal issue - it may still be useful to record the probe in Beam anyway.