2018-08-20 Meeting notes

Tim Whitlock (Deactivated)

• foghorn_leghorn (Unlicensed) should be able to progress
  Error rendering macro 'jira' : Unable to locate Jira server for this macro. It may be due to Application Link configuration.
  now Fran Benoy-Deeney has returned from holiday and can review the fixes. foghorn_leghorn (Unlicensed) will also let her know what other fixes have been done that will be included in this beta so that they all can be checked before actual release.
• A meeting about RaDaR tickets will be held when the RaDaR feeds are back working. These depend on the ukrdc fix currently being tested on staging and the medications being imported currently being re built to remove duplicates.
• George Swinnerton will chase SSG about the logging of xml to the importer log files. These are causing disk issues with the logging files now the programmers have returned from hoilday they should be able to progress this.
• foghorn_leghorn (Unlicensed) is awaiting the return file from Chloe to progress this further.
• George Swinnerton has made a start on the specification. The reason behind this document was discussed and its basic premise discussed. The document should detail how the dataset elements should be populated using the data fields in the ukrdc. In turn the schema fields that populate the fields should be documented. The specification should detail how the fields that contribute to any given dataset field should be processed to produce the out put data. This should then provide sufficient information for a programmer to implement the extract. Decisions will need to be made on what the precise calculation rules should be for some fields, what to do if data is missing etc. This document should act as the reference for testing the extract and verifying the values returned. It is likely that it will be modified as the true nature and variability between units of some ukrdc fields contents is understood. George Swinnerton hopes to have a provisional draft within two weeks 03 Sep 2018 
• rapolas (Unlicensed) brought up the problems of the current RaDaR export process. this uses a library tablib which builds the data in memory meaning as a database grows the library can no longer process all the data due to memory limitations.A discussion was had over the what options to deal with this might be. Currently the library is used to produce csv and excel files. The excel file could be produced from the csv after export but would not then be able to take advantage of colouring data cells during export as happens with the rituximab export. In addition the export process currently takes a significant time (5+minutes) meaning integrating a report button would not be as simple as click and download the report due to the delay. No obvious solution was suggested and rapolas (Unlicensed) will investigate alternative libraries (such as openpixel and pandas) since it is unlikely that a stream option will be added to the tablib library in the near future.

• rapolas (Unlicensed) is expecting to have to run Survey reports for the programmes team for the YourHealth surveys this week or early next
• George Swinnerton is expecting to deal with the infection data (PHE England Infection Data) as soon as the Data Sharing Agreement has been finalised.

As a result of a Data Manager error EPO data for Bradford,Hull, York and Leeds has not been loaded into the RenalReg database. This needs to be backloaded. foghorn_leghorn (Unlicensed) will do this when Fran is back in.