Many types of functions produce data amenable for charting in the xy dimensions. These functions are defined as
f(x[n]) = y. To transform this function into an
f(x) = y function, provide a projection that takes a set of
(x[n], y) values and picks a single
(x,y) pair. In practice the set
(x[n], y) values must be limited to fit within memory/runtime bounds... ergo for a given x there must be a limited number of x[n-1].
In summary a projection can produce xy data from x[n]y data provided:
- there is one value y for each (x[n])
- the number of x[n-1] is limited for a given x
Many types of data end up satisfying these conditions, in particular measurement data. Measurement data can be modeled as ty. If measurement data can be overridden then add a dimension to track when the data was recieved, ie tty.