Review of reconstruction parameters and defaults

[cherylepatrick]

I think it would be useful to have a review of the parameters available to be passed to the various stages of reconstruction, to ensure that

• We are clear exactly what all of these parameters mean (and they are all explained in one place)
• We know what the default values are set to and why
• Where we are not sure of the best default value for a given parameter, we should make a note and plan for additional studies to optimise them

This is probably related to the reconstruction validation - we need a standard reconstruction that we fully understand. (I'm sure that plenty of this IS already understood but just not by me. But we should get it all organised and in one place).

[drbenmorgan]

Added @pguzowski as initial assignee as this will primarily be a task for C&R.

Relevant related issues: #62 #57 #33

[drbenmorgan]

I'm in the process of developing a proposal to simplify configuration in flreconstruct and make it more robust. It won't directly address @cherylepatrick's points, but should help towards them. It's also related to requests from C&R on preparing standardised pipelines (versioning etc). Once I have a prototype, I'll post more.

[drbenmorgan]

As this is related to topics discussed at yesterday's Analysis Board meeting, I'm adding back @cherylepatrick, @emchauve and @pfranchini. They'll be preparing material for the tutorial at Bordeaux, which can contribute to this task.