Using xpnt the wrong way?

Using x2pnt(waveName, x1 ) for a wave that is, for example Param v Time, and  x1 is earlier than the first point in the wave, a negative number is returned.  Mutiply this number by the deltax for the wave, and the answer appears to correspond roughly to how much earlier  x1 is from the first point in the wave.

Is this negative returned number a hidden feature?


Yes, hidden. Internally, it just does what we recommend in our documentation: round((x-x0)/dx) without trying to clip or throw an error. It will probably remain undocumented, which means we are more or less free to change it if that becomes somehow compelling. But I think that's a pretty remote possibility.

No, please don't remove the useful (to those who know of it :-) feature.   Useful for extrapolation.