Should we remove father selection from crossColonies()?

[janaobsteter]

Currently, we have a duplicated situtation:

• crossVirginQueens can cross multiple virgin queens at once and can only select fathers for each of them from provided drones

• crossColonies duplicated the code for selecting fathers from the provided drones.

  fatherGroups <- pullDroneGroupsFromDCA(
    DCA = drones,
    n = nCol,
    nFathers = nFathers,
    removeFathers = removeFathers
  )

  and then

  for (colony in seq_len(nCol)) {
    ret[[colony]] <- crossColony(
      colony = colonies[[colony]],
      drones = fatherGroups[[colony]],
      nFathers = nInd(fatherGroups[[colony]]),
      removeFathers = removeFathers,
      checkMating = checkMating,
      simParamBee = simParamBee
    )
  }

  So I think, we should use crossVirginQueens to cross them and then only put queens in the colonies.

[gregorgorjanc]

This will likely happen when we merge L2 and L3, no?

Still, I am not sure, what is the real problem? That this code is duplicated

 fatherGroups <- pullDroneGroupsFromDCA(
      DCA = drones,
      n = nCol,
      nFathers = nFathers,
      removeFathers = removeFathers
    )

I think that this is controlled by nFathers - there is no sub-sampling when nFathers == nInd(fatherGroup), right?

[janaobsteter]

Should we start merging the functions?

[gregorgorjanc]

I think this would make sense. Which function do you want to start with? I suggest we do one at a time to make life simple and controlable.

[janaobsteter]

I can start with a "non-crucial", such as collapse (if it's not already done) or split.

[janaobsteter]

Ok, I now realized what the problem is. Father selection should in reality be pulled out of crossVirginQueen() - in there, we use only "simple" selectInd(). That should be taken out and instead, cross() function should use pullDroneGroupsFromDCA(), because that one has all the checks and functionality (such as "killing" used drones).

[gregorgorjanc]

Closing as this has been addressed in https://github.com/HighlanderLab/SIMplyBee/pull/316