Homepage of Martinus Werts

First-person smoother

Smoothing, when applied with care and avec discernement, can bring out specific features in noisy data, and make their presentation clearer, as most scientists know. Being a chemist, I was fond of the Savitsky-Golay smoother (see doi:10.1021/ac60214a047, doi:10.1109/MSP.2011.941097, and wikipedia).

Since reading the paper "A perfect smoother" by P. H. C. Eilers (doi:10.1021/ac034173t), and trying it out (with help from undergraduate student Simon Bordeyne), I now prefer to use the Whittaker smoother for noisy data! See for example our recent ChemPhysChem paper.

To celebrate this, I have taken the time to put my Python version of the Whittaker-Eilers smoother on GitHub. Use at your own risk and for the good of mankind.

Homepage of Martinus Werts

Other articles

A photograph of the Boltzmann distribution

Ph.D. scholarship (2016-2019)

Python implementation of the Fraser-Suzuki 'skewed Gaussian' function

Projets pour mécatroniciens