See also
Octave plots
The figure below is from (Edgar and Flyvbjerg 2018). All histograms were generated from the same simulated log-normal abundance distribution. The histograms in a given row are generated from an identical incomplete enumeration of this distribution, i.e. a set of simulated reads which does not capture all OTUs in the sample. From top to bottom, the enumerations contain 8k, 4k, 2k and 512 reads, respectively. Binning rules are: Octave (Edgar and Flyvbjerg 2018), Williamson (Williamson and Gaston, 2005), Magurran (Magurran, 2004) and Preston (Preston, 1948). Shape distortions are circled in red. Octave binning is the only method that preserves the shape without misleading distortions. Bins are labeled 1, 2 ... 9 in order of increasing abundance.
