General estimates will be given for reconstructing a function from sampling it at a discrete set of points. Noise will be introduced and exponential probability results are exploited. A measure of data is introduced which provides a dichotomy, rich data versus poor data.
Copyright. All Rights Reserved.