Re: [Jsbsim-devel] FGTable breakpoint data check

[Jon S. B. <jon...@co...>]

> I'll take a different stab:  while there are two independent variables
> (alpha and flap position) given, this is only a one dimensional (i.e.
> one input) table!  So the table is being read in (somehow) as a 2D
> table and then going nuts from there.
> 
> Commenting out the flap position independent variable lets the check
> progress to the error Ron found.  Correcting that one as I suggested
> sends us to the next error, CDDe which is truly not monotonically
> increasing:
> 
> Note: It doesn't look like Andreas monotonicity check is incorrect
> after all.  Only that the data provided wasn't sufficient and things
> went bad after that.  Maybe it points to a need for a check to ensure
> at least three columns of data when there are two independent
> variables.
> 
> Patch for SGS.xml is attached.
> 
> Dennis

Good catch, Dennis - and, yes, there should be another check to make sure
the data and table definition agree.

This is a larger discussion than I have time for this morning. Can someone
summarize it? Is there anything other than a data point change and sanity
check on the table that needs to be added?

Jon