Direct loading data from databroker. ROI settings tab. #191
Conversation
…tabroker is added
There was a problem hiding this comment.
Thanks for the update. I'd like to mention, that it's not possible to give a comprehensive review on such a lengthy PR with 60+ commits, more than a thousand line changes in 14 files. Each bullet point in the description deserves its own separate PR.
Anyway, here is my review.
Co-Authored-By: Maksim Rakitin <mrakitin@users.noreply.github.com>
Co-Authored-By: Maksim Rakitin <mrakitin@users.noreply.github.com>
Co-Authored-By: Maksim Rakitin <mrakitin@users.noreply.github.com>
Co-Authored-By: Maksim Rakitin <mrakitin@users.noreply.github.com>
Co-Authored-By: Maksim Rakitin <mrakitin@users.noreply.github.com>
Co-Authored-By: Maksim Rakitin <mrakitin@users.noreply.github.com>
Co-Authored-By: Maksim Rakitin <mrakitin@users.noreply.github.com>
Co-Authored-By: Maksim Rakitin <mrakitin@users.noreply.github.com>
Co-Authored-By: Maksim Rakitin <mrakitin@users.noreply.github.com>
| @@ -181,18 +313,10 @@ def update_roi(self, element_list, std_ratio=4): | |||
| self.roi_dict.update({v: roi}) | |||
|
|
|||
| # remove old items not included in element_list | |||
| for k in six.iterkeys(self.roi_dict): | |||
| for k in self.roi_dict.copy().keys(): | |||
There was a problem hiding this comment.
This is one of many things in the original code that did not work as expected.
pyxrf/model/load_data_from_db.py
Outdated
| Non-existing scans in the range or scans causing errors during conversion will be skipped. | ||
|
|
||
| Keyword parameters | ||
| ------------------ |
There was a problem hiding this comment.
In particular, I meant a separate Keyword parameters block. I think everything can be under the Parameters above.
pyxrf/model/load_data_from_db.py
Outdated
| num_end_lines_excluded=None): | ||
| """ | ||
| Transfer multiple h5 files. | ||
| Load data from database and save it in HDF5 files. | ||
|
|
||
| Parameters | ||
| --------- |
There was a problem hiding this comment.
| --------- | |
| ---------- |
pyxrf/model/load_data_from_db.py
Outdated
| break | ||
| except Exception: | ||
| # Wait for 10 minutes and retry | ||
| time.sleep(600) |
There was a problem hiding this comment.
Can it be a configurable parameter? Otherwise a new version of the code will be needed each time a new sleep time is asked.
Implemented and fixed features include:
-- direct loading data from databroker by scan ID is working. Remotely tested on SRX, HXN and TES.
-- improved consistency of behavior of
make_hdfandpyxrf_batchfunctions. Both functions will raise exception if error occurred during the processing if called for a single scan/file or go to processing the next scan/file if the list of files or range of scan IDs is provided.make_hdfwill also may detect 'incomplete' scans that have no stop document.-- fixed and improved functionality of
ROI settingtab, which allows to compute ROIs. All features are operational and stable. Added an option to subtract background from raw data before computing ROIs.-- fixed minor stability issues, including program crash when removing all selected lines using
Delete not selected itemsorDelete Rel Intbuttons inFittab.-- working directory is now set at start-up of PyXRF to current working directory.
-- calculated ROIs are saved in the HDF5 files. Separate set of ROIs are saved for each dataset (such as
sum,det1,det2etc.-- all processing results contained in HDF5 file are always loaded. For example, if fitting or ROI results for
det2are saved in an HDF5 file, they will be loaded and available for viewing even if raw data from individual detector channels is not loaded (onlysumis loaded).-- buttons
<and>are added toFitandSpectrum Viewtab to switch between element emission lines (possible fix for issue with MacOS version of PyXRF, in which ComboBox doesn't seem to keep focus and swiching using Up and Down arrows does not work reliably).