The Big Run Status Page - UVA-CAMA/NICUHDF5Viewer GitHub Wiki
"The 2nd Big Run" is complete! The sites will be tasked with using BAP v2.1 and MergeTags v2.1 on all of their data.
-
Issue recorded: There is 1 hour off in Miami UFC collected data. it might be 25 hours or 23 hours for some infant days.
-
Speed test (UVA 10 infants)
-
BAP 2.0 ~22sec per daily file and total of 9mins for 30 days of data
-
BAP 2. 1 with HCTSA (overwrite previous BAP2.0 result) ~17sec per daily file and total of 3mins of 10 days of data
-
BAP 2. 1 without HCTSA (overwrite previous BAP2.0 results) ~15sec per daily file and total of 3mins of 10 days of data
-
BAP 2.1 with HCTSA (no previous BAP2.0 result) ~29sec per daily file and total of 30mins of 100 days of
-
BAP 2.1 without HCTSA (no previous BAP2.0 result) ~26sec per daily file and total of 17mins of 50 days of data
-
Here is the 2nd big run status as known to the LDCC:
| Site | CWRU | NU | WUSTL | UAB | UM | Total |
|---|---|---|---|---|---|---|
| Received Email | 9/30/21 | 9/30/21 | 9/30/21 | 9/30/21 | 9/30/21 | |
| UFC 1.5.0 | N/A | Complete | Complete | Complete | Complete | |
| UFC 1.6.0 | N/A | n% medi per infant | 30% medi | |||
| BAP 2.2 | Complete | Complete | Complete | Complete | Complete | |
| MergeTags 2.2 all | 100/100 | 136/138 | 59/60 | 265/265 | 163/176 | 723/739 |
| MergeTags 2.2 40w | 98/98 | 133/133 | 54/54 | 259/259 | 163/173 | 707/717 |
| Final Status | Complete! | Complete! | Complete! | Complete! | Complete! | |
| Total Expected | 98/100 | 133/138 | 59 | 259/265 | 173/176 | 717/739 |
| Issue reported | 99 logmerged | Files 4103, 4112, 4134, 4139, 4152, 4154, 4155, 4156, 4060 are empty files (An inadvertent ~8 week gap in the DWC recorded data);UFC conversion failure: 4172. |
"The Big Run" is complete! The sites have been tasked with using BAP v2.0 and MergeTags v2.0 on all of their data.
Here is the status as known to the LDCC:
| Site | CWRU | NU | WUSTL | UAB | UM | Total |
|---|---|---|---|---|---|---|
| Received Email | 3/31/21 | 3/31/21 | 4/2/21 | 4/1/21 | 4/2/21 | |
| Files Converted | N/A | Complete | Complete | Complete | Complete | |
| BAP | Complete | Complete | Complete | Complete | Complete | |
| Results Merged | Complete | 136 sent | Complete | Complete | 161 sent | 721 sent |
| 10 HDF5 Files | 4/2/21 | 4/7/21 | 3/5/21 (not @30 DOA) | 4/3/21 | 4/23/21 | |
| Logs Checked | 100 | 136 | 59 | 265 | 161 | 721 checked |
| Final Status | Complete! | Complete! | Complete! | Complete! | ||
| Total Expected | 100 | 137 | 59 | 265 | 176 | 737 |
| Issue reported | added filter | Excluded from study:2008 | An inadvertent ~8 week gap in the DWC recorded data: 4103, 4112, 4134, 4139, 4152, 4154, 4155; UFC conversion failure: 4167, 4169, 4172, (4175 small); Exclude from study: 4042, 4053, 4054, 4060, 4156; | |||
CWRU
- SPEED TEST: Julie ran one infant with 89 twelve hr files (~44 days of data).
- Took ~25sec per file and total of 37min to run
- CPU time ranged from 15-33%
- memory increased from 27% before program to ~31% during with 640-1400MB
- Need to re-run with filtering
NU
- About 40% of the cohort uses medi files, which will need to be reconverted
- The rest of the files are stp files, which will not need to be reconverted
- Needed command line version of BAP - created it and posted it here: https://github.com/amedwards/NICUHDF5Viewer/wiki/BAP-Command-Line
- Issue in medi files: https://github.com/UVA-CAMA/NICUHDF5Viewer/issues/28 - medi files were not converted with newest version of file converter (they were converted with 1.1.7). NU must have grabbed the wrong files - they are looking into it.
- Issue in stp files: https://github.com/UVA-CAMA/NICUHDF5Viewer/issues/25 - same as issue in WUSTL files with redundant timestamps. Issue is rare, so will not fix.
WUSTL
- Explanation of the (non)issue with hdf5 files from stp: https://github.com/UVA-CAMA/NICUHDF5Viewer/issues/25
- Done!
UAB
- Done!
UM
- Need to export from Philips data warehouse a good number of the recently enrolled patients and some are still<40 w
- medi files will need to be reconverted
- Converting and running BAP in parallel
- Apnea algorithm glitch determined to not be a problem, BAP run may continue. See error details here https://github.com/amedwards/NICUHDF5Viewer/issues/20
- SpO2 signal occasionally missing: patch in next release: https://github.com/UVA-CAMA/NICUHDF5Viewer/issues/27
LDCC
- Transferred file location:
- Apnea files for Karen to review are located here: CAMA\PreVent\ApneaValidationKaren
- ResultsMerged and LogMerged files are here: CAMA\PreVent\MergedResults_BAP2_0