People
Documentation
Notes
- Impromptu Modelling Meeting: To Talk About Getting the Model Going
- No computers were allowed to begin with
- On discussing what needs to be done it became clear that Mark and Andy had different ideas about what was going on!
- Mark was under the impression that Version 3 of the population initialisation was to operate only on households alone, whereas, Andy thought it was still to deal with Communal Establishment (CE) populations
- Andy and Mark scrambled for documentation
- Andy produced a print out of Andy Turner's MoSeS Meeting 2006-05-23 Page Version 0.1.1
- Mark emailed a copy of 2006-05-23MBDocument.doc
- It was made clear that Mark's document was the one that detailed what Version 3 was to be
- So, how to operationalise it?
- How to control (constrain) and optimise for household population counts using CAS data?
- Upon re-inspection of the Census CAS data it again became clear that handling CE populations was probably the most appropriate way forward - as most of the variables wanted to be used in optimisation are only available for population counts (which include CE populations), not household population counts
- Mark produced a copy of 2006-05-23MBDocument.doc which he had edited during the meeting on 2006-05-23 (laterly this was emailed on request and is stored as Mark Birkin's Population Reconstruction Model Word Document 2006-06-16)
- It was decided that Andy would continue with what he had envisaged. The next tasks ahead are as follows:
- The aggregated MSOA tables are to be merged into a single table
- A convenience method written so that households populations (sets of HSAR records) for HRPs of specific age can be obtained easily
- A data handler written for CAS003 so that age profile optimisation can be done for age breakdowns of household and CE populations
- Optimisation and control constraints for age need re-working to take account of the difference in age grouping in the HSAR
- It is still uncertain how long the program will take to develop and run
- Significant development of code should be made in the next week.
- It is not urgent to develop results for 2006-07-13 Meeting with Mike Davidge and Mike Robinson at East Leeds PCT. This meeting can be based on other work.
- To wow Mike Robinson et al we really need a map of estimated LLTI change or something health related like that for 5-10 years hence. Clearly this requires not only the initialisation of the population, but also the dynamic simulation.
Actions
- Andy is to report on progress by email on 2006-06-22
- Andy is to communicate with Mark if any major problems are encountered in the mean time