Release Candidate: Experimentation Orchestration

convenience methods I wrote for the nonlinear examples in the paper. These basically take a wrapper for a mud problem and run experiments with increasing numbers of measurements, then generate the summary statistics for the errors.