Why do integrate_to_ppi VIR values significantly change between bioRad v0.5.2.9410 -> v0.6.1? about biorad

[bart1]

@BerendWijers as we already figured out quite a bit off the difference (relating to approx fun interpolating over NA's, and the shift in the cell centers iirc). I don't remember if those only resulted in differences in R and overlap or also partially VID? it would be good to already report that here. That saves a lot of double work.

[BerendWijers]

Hi @bart1 good call - I'll add that information too.

I notice this is a rather specific question. I'll dig into this myself and report what I find.

Can this be assigned to me?

[BerendWijers]

Hi all,

@bart1 and me dug deeper and we've found the cause. I'm closing this issue and I've made a report about integrate_profile

Small numerical differences arose from the following functions

• scan_convert.R::scan_to_spatial
  • uses cell centers now

Major differences:

• integrate_profile
  Closing issue as this is not directly related to the function integrate_to_ppi, see #543