-
Notifications
You must be signed in to change notification settings - Fork 10
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
Implement the Correlator special suite #328
base: dev
Are you sure you want to change the base?
Conversation
This pull request introduces 1 alert when merging 5dbc0f1 into 2331f1d - view on LGTM.com new alerts:
|
32d2fc5
to
e5be471
Compare
This pull request introduces 1 alert when merging e5be471 into 2331f1d - view on LGTM.com new alerts:
|
Hmm, it's failing because the CI can't install |
e5be471
to
4288618
Compare
This pull request introduces 1 alert when merging 4288618 into f8d225d - view on LGTM.com new alerts:
|
metadata for plotting. | ||
""" | ||
xr_attrs = { | ||
"xlabel": xlabel, |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
When receiving data as an xarray.DataArray
(as you create here), there is already a documented protocol in metropc
to use the coordinate label(s) as X axis label in 1D, whereas the y label is passed as an attribute (like here).
Would you mind using this instead of an xlabel
attribute for compatibility?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Yeah sure, that should be fine.
4288618
to
c27cd0c
Compare
This pull request introduces 1 alert when merging c27cd0c into 1489977 - view on LGTM.com new alerts:
|
c27cd0c
to
46086ae
Compare
This pull request introduces 1 alert when merging 46086ae into 86177ea - view on LGTM.com new alerts:
|
46086ae
to
e685bbe
Compare
This pull request introduces 2 alerts when merging e685bbe into 621f1e5 - view on LGTM.com new alerts:
|
This pull request introduces 2 alerts when merging 2759386 into 7b704c3 - view on LGTM.com new alerts:
|
2759386
to
4b8f0e0
Compare
This pull request introduces 4 alerts when merging 4b8f0e0 into 56f56f5 - view on LGTM.com new alerts:
|
This is to match what the DataTransformer does.
This makes it easier to change the settings from within special suites, without having to dig into _SpecialAnalysisBase code.
This pull request introduces 4 alerts when merging 3a135f8 into b9e30b5 - view on LGTM.com new alerts:
|
This is a very simple implementation of something that can hopefully be improved and cleaned up in the future: support for context-defined annotation widgets. Each ROI is defined as a parameter, they are added to a View with an argument to the Image_WithRoi type, which is read by the Correlator and thus creates an ROI. When the ROI is changed, an updated parameter is sent to metropc and the Correlator modifies the context file with the new ROI parameters (position, height, etc).
Also added a warning dialog that appears when the user tries to move a ROI when there's a syntax error in the code (which libcst will fail to parse).
Among other things, this makes it possible to downsample all series being plotted.
This is to allow LinearROIs on images.
This is in preparation for using another library that's most up-to-date on conda.
- Replace np.warnings with the warnings module - Replace np.float with float - Pass enum values instead of the enums themselves to redis functions - Drop support for saving float arrays PNG images and expecting them to be restored with the same values. Seems that Pillow dropped support for it.
Lots of minor fixes to the tests, and fixed one bug in the correlator that caused an extra nan value to be added to the error arrays for each series.
We should probably aim to point this at |
I think you pressed the wrong button there 😄 |
I did, apologies 😳 |
Finally got around to making a PR out of this 😅 I'm keeping it as a draft for now because there are a couple of (hopefully) minor issues still to fix:
high_high_water_mark
branch ofmetropc
, which in turn is based onfeat/bound_device_features
. Thehigh_high_water_mark
branch just increases the HWM for the output socket ofmetropc
, because I noticed that sometimesmetropc
was just dropping outputs and that fixed the issue. It could also be because the special suite is not reading fast enough, but that needs to be investigated. And perhaps we should wait to merge this until thefeat/bound_device_features
branch is merged too.To reviewers, even though this is a draft I would suggest starting now because there's... a lot... of code. There's also some auxiliary commits that were needed for the Correlator, so I would suggest reviewing each commit individually because they're fairly atomic (it'll make more sense that way).
Note: this PR depends on #326, so you should ignore commits 1581c1b and 479e93a.