2016 05 23_meetingsummary

Morning meeting

Present: Robert Clegg, Bastiaan Cockx

Bas's visit to Newcastle

• Very positive, people there keen to use iDyno2
• IBM in microbiology wiki, with iDyno2 as a sub-wiki?

Output as XML

• Bas has finished the structure, but is still implementing it throughout the source code
• Currently working on Reaction stoichiometry
• Shape classes are done, but not Dimension
• Much of the stuff in ObjectFactory is being replaced with NodeConstructer approach
• Bas is uncertain about how to read/write concentration arrays. Rob will check this

Removing aspect interface from process managers?

• Leave for now, but needs to be done using XML output/input.
• Obligatory aspects, requirements?

Reaction Library

• Use as default, allow extra/exclude in a Compartment
• Should be easy to code, as follows the SpeciesLib approach
• Solutes could also be done, needs more thought

Process managers

• SolveDiffusionTransient: Need to clarify how the agents grow
• SolveChemostat: Need to do agent growth!

Boundaries

• Rob has been working on Agent and soute boundary methods, getting there!
• Tricky defining what is attached (i.e. biofilm) and what is floating in boundary layer
  • Boolean flag for sloughing on boundary layer
• Poorly defined boundaries specifying solute concentrations on adjacent boundaries?
  • Spatially structured compartment with well-mixed sub-regions
  • Could we prohibit this?
  • Or flag a message and make not well-mixed?

Agent events for Toledo group

• run other events?
• change species (death, etc)

Afternoon meeting

Present: Robert Clegg, Stefan Lang

Restructure from Grid to Shape

• Rob started restructure, Stefan has been continuing it
• Making great progress but not quite done yet: errors
• Resolution calculator: scale resolution instead of setting length as arc length - Rob agrees
• Initialise with UniformResolution by default.. should we set this in the XML?
• Enter into XML as degrees? Or read in "pi" string? Should be sorted when we introduce units
• Grid types: strip away old grids and make SpatialGrid non-abstract
• Iterator has switched from Grid to Shape elsewhere in code
• Move stuff in linearAlgebra.PolarArray into CylindricalShape and SphericalShape
• Stefan will let Rob know if any of the checkDimensions methods in Matrix or Array are causing problems, so he can fix it.
• Rob will check Stefan's new approach to storing current boundary in CartesianShape
• Setting the number of voxels per shell/ring is still unclear, especially the N_ZERO_FACTOR stuff. Stefan will clarify what is going on here!

Manuscript

• Design concepts: ODD uses grid cell to mean many things, we should use "voxel"
• Sub-voxel points: shape.subvoxel package and SolveDiffusionTransient ~lines 150-250