New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
dials.import: individually select models from reference expt #1371
Conversation
Codecov Report
@@ Coverage Diff @@
## master #1371 +/- ##
==========================================
+ Coverage 65.72% 65.73% +0.01%
==========================================
Files 615 615
Lines 69122 69151 +29
Branches 9550 9553 +3
==========================================
+ Hits 45429 45459 +30
Misses 21854 21854
+ Partials 1839 1838 -1 |
We discussed this over at dials-east yesterday - the idea is fine but we felt that the proposed command-line API was cumbersome. A suggestion was to use
or
as the default - by analogy with fixing parameters in refinement. However it's also noted that there is a lot of inconsistency across dials with how command line options like this are presented. @dwpaley how would you feel about this alternative command line scheme? |
Probably the place most likely to encounter this pattern is in refinement where it's used to turn on/off refinement parameters e.g.
|
I think that sounds great. Will have to rewrite the test and get back to you. |
511a407
to
23cedfe
Compare
Thanks, @graeme-winter and @ndevenish. It's now a single phil entry using type=choice(multi=True) as in Nick's example. What do you think? |
So, I know we sort of asked for it, but I'm really not sure how well suited the So, given that this parameter is probably expected to be used for turning on/off single I'd lean towards either the original solution (sorry!) or something custom that just takes a I'm leaning towards the original multiple parameters, which is a bit of a mouthful but easy to understand and explicit. Thoughts? |
Hi Nick, thanks for your comments. I agree that your example is confusing, and that clunky + explicit is better. I think the first commit is still ok if we just get rid of the others. Is that what you were thinking? |
Yes, I would be fine with that. It's explicit enough that we could always add a combined parameter later, if we were really eager. @graeme-winter fine with this? |
Add new phil parameters input.use_beam_reference, use_gonio_reference, and use_detector_reference. If set to false, the corresponding models are not imported from the reference geometry file.