ABC

For my inference-of-variances project, I got ABC working. It is delivering a correct posterior, even at a reasonable (finite) distance threshold. All is good! I put a figure showing this into the nascent paper.