/
2019-07-15 Meeting notes

2019-07-15 Meeting notes

Date

Attendees

Goals

  • Review current task status and tasks for the week. 

Discussion items

ItemWhoNotes
Outstanding Tasks
  • Kings have started sending some of the data George has asked for. There is a greater range of of modalities, Home HD patients data indicates if home or satellite. However treatment locations have been received using their internal names and not our names. This has shown a flaw in the RDA_to_RDA code which doesn't  validate these codes so they end up in the database and only throw an error when validating the extract produced from the database. George Swinnertonwill fix this issue which will then ensure only valid satellite codes are saved. Test results are only from 2018 so using the data to derive start dates that are earlier is not currently possible. They have agreed to try sending the full history for one patient to ensure the large file doesn't cause any other issues. We will also have to look at the processing of multiple files and whether we will need to use a big file channel that runs a single process as has been implemented on the PV channels.
  • The extract codes is more or less complete there is still some work to do on the blood tests associated with dialysis sessions and kings need to send the full list of test results as currently they are limited to those required for PV extracts not the audit extract. 
  • George Swinnerton highlighted that the addresses might need re-validating due to the addresses now being split across a different number of lines and so appearing as if they have changed compared to the previous multi-line structure. Due to QAS data files expiring on Fran's laptop she wasn't able to test this. George Swinnerton will test this locally.
  • The versioning is now completed on the Schemas for UKRDC and PV but there is still some work to do on the validation schemas.@
  • George Swinnerton has tested the Studygroup memberships and it works as expected. It needs some documentation to clarify it however it works as expected and can be used to group data by study group with a suitable ID.
  • rapolas (Unlicensed) has completed the documentation for the  UKRR Extract Code (https://confluence.ukrdc.org/display/TNG/UKRR+Extract). Andrew Atterton , james.griffin (Unlicensed) and George Swinnerton to review and feed back if any more details required. George Swinnerton to check if he has made changes that need extra documentation.
  • rapolas (Unlicensed) has documentated most things. he still needs to document the transfer processes on rr-hs-test.
  • Sheffield have fixed their issue and Stoke have been given similar information as they were also sending an incorrect test code.
  • George Swinnerton has done some work on the validation ticket RR-1316 which is due to be released out side of the sprint. No work log has been filled out on the ticket so will need back filling. Andrew Atterton has done a ticket but has found issues with black and using it on python 2.7. He is currently upgrading the code base to python3 which will remove the black issue and will be required by the new year.
  • rapolas (Unlicensed) is currently tracing all RaDaR patients. George Swinnerton flagged Fiona had been saying medication data needed to be imported from HSMig which already happens. George Swinnerton had found a patient where not all the data was there. rapolas (Unlicensed) will look into it as it suggests a syncing issue.
 Versioning
The current version process is to use x.x.x structure however there is a move to use a date based structure for software releases making it immediately obvious what the age of the code is. This type of version could be useful for the UKRDC and validation where users report issues as it would be quickly clear if the user had the latest version or not. How easy a change would be needs to be considered. Also whether there could be implications. Due to the way validation changes are made the version is applied to several parts of the code. Whilst in principal it was felt that the versioning could help how it could be implemented needs to be considered. George Swinnerton pointed out there is some devpi library crossover between the UKRDC and validation and due to the way the current validation code base is built some of these would appear to be updated even when they haven't.  ALl will go away and think about the implications when working on the code to see what would be the best way forward. rapolas (Unlicensed) pointed out that using a changelog would make the current versioning process better however even where one is in use it is not always being updated like the current ukrdc changelog. 

Action items

  • George Swinnertonto fix the RDA_to_RDA validation code to validate the satellite codes  
  • George Swinnerton to document the study group memberships and how to set them up and extract data for a given study group.  
  • Andrew Attertonjames.griffin (Unlicensed) to review the UKRR Extract code documentation and identify any extra info they would benefit from.  
  • George Swinnerton to review the UKRR Extract code documentation and identfiy any exra information that should be added and if changes he has made need extra documentation.  
  • rapolas (Unlicensed) to add documentation on the transfer process on rr-hs-test