scans showing error, no dimensional information

I am not using find nominals it is set to nominals, not find noms

Are they DEFINED or RELEARN scans?  Defined scans follow the path that was generated and will return zero data for areas where the probe either leaves the surface or where the deflection is too high.  Relearn scans will attempt to adjust themselves to track the surface so that the probe deflection always stays within acceptable limits.

When it comes to DEFINED or RELEARN. Is it up to the programmer i.e. me. Or is there a method you recommend more over the other? It sounds like these are similar methods. So relearn will be more likely to give you all the hits you ask for and adjust for probe deflection, whereas DEFINED will ignore data due to probe deflection?

It's up to the programmer to choose the appropriate scan method for their application.  Defined scans are generally better for application where you don't expect to encounter a large amount of deviation from nominal - machined parts with less than 1mm profile tolerances to fully constrained datum reference frames for example .  Relearn scans are better for parts which are expected to deviate significantly from their nominal shape or where the location of the surface being scanned can vary.  However, because the scan essentially teaches itself as it goes, they need to run at slower scan speeds than a defined scan.

It's up to the programmer to choose the appropriate scan method for their application.  Defined scans are generally better for application where you don't expect to encounter a large amount of deviation from nominal - machined parts with less than 1mm profile tolerances to fully constrained datum reference frames for example .  Relearn scans are better for parts which are expected to deviate significantly from their nominal shape or where the location of the surface being scanned can vary.  However, because the scan essentially teaches itself as it goes, they need to run at slower scan speeds than a defined scan.