You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
As per latest commit (ec5e475...) in jungfrau_flex_geom, a geometry file is written,
and the tile-wise blocks incl. corner_x/y look reasonable.
The problem is that this is still incomplete information for use with JUNGFRAU_geom.from_crystfel_geom('test2.geom')
I use the geometry base class for the writing, as there are no detector-specific inheritances yet. This implies, as I see it, a problem, because the base class requirements are likely too close to the AGIPD properties.
For instance, the concept of quadrants makes less sense, at least from the general point of view (1, 2 and also the 6 SPB JF-units are not compatible with this).
In order to skip refactoring considerations etc. for the moment and just provide a geom file with formally "correct" (?) content, I complemented the auto-generated geom file with quadrant and module rigid group definitions. But this still gives rise to an error, so I'm still missing something ...
The text was updated successfully, but these errors were encountered:
I'd suggest we get the JUNGFRAU geometry working and merged first, and worry about reading/writing geometry files in a separate PR. You can raise NotImplementedError for anything that's known to be broken at the moment.
I'd guess that the rigid group definitions will need to be generalised. We should probably define only each module as a rigid group, ignoring 'quadrants' for JUNGFRAU.
As per latest commit (ec5e475...) in jungfrau_flex_geom, a geometry file is written,
and the tile-wise blocks incl. corner_x/y look reasonable.
The problem is that this is still incomplete information for use with JUNGFRAU_geom.from_crystfel_geom('test2.geom')
I use the geometry base class for the writing, as there are no detector-specific inheritances yet. This implies, as I see it, a problem, because the base class requirements are likely too close to the AGIPD properties.
For instance, the concept of quadrants makes less sense, at least from the general point of view (1, 2 and also the 6 SPB JF-units are not compatible with this).
In order to skip refactoring considerations etc. for the moment and just provide a geom file with formally "correct" (?) content, I complemented the auto-generated geom file with quadrant and module rigid group definitions. But this still gives rise to an error, so I'm still missing something ...
The text was updated successfully, but these errors were encountered: