skbio.diversity.alpha.robbins

skbio.diversity.alpha.robbins(counts)[source]

Calculate Robbins’ estimator for the probability of unobserved outcomes.

State: Experimental as of 0.4.0.

Robbins’ estimator is defined as:

\[\frac{F_1}{n+1}\]

where \(F_1\) is the number of singleton OTUs.

Parameters

counts (1-D array_like, int) – Vector of counts.

Returns

Robbins’ estimate.

Return type

double

Notes

Robbins’ estimator is defined in 1. The estimate computed here is for \(n-1\) counts, i.e. the x-axis is off by 1.

References

1

Robbins, H. E (1968). Ann. of Stats. Vol 36, pp. 256-257.