Support reading noise data with smooth transitions from point to point

[oubiwann]

If the data generated by loise is to be used by other applications, and there is an expectation for smooth transitions from point to point, there are several ways to provide this:

1. Follow gradients - see Add gradient functions #36
   • at peaks or valleys, randomly choose a new direction
   • follow that gradient
   • this may require special logic to escape local minima/maxima as well as repeatedly following the same gradient paths
   • this is now scheduled for 1.4.0
2. Follow a random walk - see Implement a random walk scan through noise data #53
   • just need to be aware of boundaries and implement a random "bounce" from these
3. Follow a boustrophedon - see Add "linear scan" function #54
   • start at the beginning of the data
   • when the max width is hit, starting reading the adjacent point, one row down
   • when the min width is hit, go down another row, and repeat

[oubiwann]

These all have tickets now, so I'll close this.