Fix definition of Hexp

[zkamvar]

Currently, Hexp, produced by the function poppr is cited as Nei's expected heterozygosity (Nei 1978). Equation 2 in this paper translates to the implementation here if you consider x to be a genotype instead of an allele. How this came to be is a mystery. This is actually an unbiased measure of genotypic diversity representing the probability of drawing two genotypes from a population and having them be different. It's implementation in the function locus_table is accurate.

Tasks:

• Change citation
• Add average expected heterozygosity (equation 3 in Nei 1978) to poppr output
• Fix equation in "algo" reference manual.

[zkamvar]

This has been fixed in 9eaf52c

Now all that's left is to fix the equation in the reference manual.