The Interpolated and Curve-Fitted Rates family is, like the families of the Stepped Rate Projections and Constant Rate Projections categories, concerned with projecting amounts over time. Instead of having the rate change smoothly and progressively over time (constant rate), or with sharp staircase-like step-changes, the rates that are input are simply data points. There are two ways you can establish the rates between the data points that correspond to the two principal functions in this family: you can interpolate or you can curve-fit.

In terms of application, ramped rates are good for things like sales or production projections, where you want the ability effectively to 'draw a curve' of the most likely scenario (stepped rates are good for contractual things like rent and loans).

• InterpedRate treats the rates input as exact rates at the points in time to which they refer. Inbetween the data is interpolated, either in the commonly used straight line (linear) fashion, or by using methods such as cubic or hermite.
  
  
• FittedRate treats the rates input as sampled data points, to which a polynomial curve is to be fitted (although you can choose linear interpolation as well). The difference between this and InterpedRate is that there is no guarantee (except with linear interpolation), that the fitted curve will run through the points. So if you want your input rates to be honoured exactly, you probably need InterpedRate, whereas if you want a smooth curve, even that means missing a few data points, you probably want InterpedRate.
  
  
• Ramp is the same as FittedRate and simply the another name for this function.