CNoise: Estimating Computational Noise in Numerical Simulations

This repository contains source code and updates from the codes originally shared at https://www.mcs.anl.gov/~wild/cnoise/.

These codes are based on the supplemental information for the papers:

• [1] Estimating Computational Noise by J.J. Moré and S.M. Wild. SIAM J. Scientific Computing, 33(3):1292-1314, 2011. doi:10.1137/100786125; PDF from earlier preprint
• [2] Estimating Derivatives of Noisy Simulations by J.J. Moré and S.M. Wild. ACM Transactions on Mathematical Software, 38(3):19:1-19:21, 2012. doi:10.1145/2168773.2168777; PDF from earlier preprint
• [3] Do You Trust Derivatives or Differences? by J.J. Moré and S.M. Wild. J. Computational Physics, 273:268-277, 2014. doi:10.1016/j.jcp.2014.04.056; PDF from earlier preprint

Estimating Computational Noise

The following information (used in [1]) is provided to encourage the estimation of computational noise in applications and to determine strengths and limitations of the code. The following scripts produce basic noise level estimates from data.

Sample Problems

Examples

• difftable Matlab/Octave: Code for generating the example difference table in Table 3.1 of [1].