use individual tables when writing out vcf files

[petrelharp]

Currently write_vcf has a ploidy argument, and will assume that adjacent samples belong to the same individual. This should be deprecated after we provide a make_individuals method, and write_vcf should have an argument that toggles whether to write individuals or genomes (maybe: output either monoploid by default or phased or `unphased).

[jeromekelleher]

This will take a bit more time than I have right now, so kicking it down the road to 0.6.2.

[petrelharp]

Also we should have an option to output a diploid genotype matrix, summing together the rows of genotype matrix for the given individuals' nodes.

[jeromekelleher]

Also we should have an option to output a diploid genotype matrix, summing together the rows of genotype matrix for the given individuals' nodes.

So we output unphased "dosage" values?

[petrelharp]

So we output unphased "dosage" values?

Right. Well, we should output diploid genotypes based on individuals in a way that works well with scikit-allel.

[jeromekelleher]

I'm going to close this after #73 is merged as this uses the individual table to generate the diploid/whateverploid individuals. We should open more specific issues for the other stuff raised here if we want them.

[jeromekelleher]

Closed in #286.