On this page I present a detailed examination of the amount of error of the representations of all ratios of odd-factors from 3 to 11 (an arbitrarily selected stopping point, commensurate with Partch's 11-odd-limit tonality diamond), for several various EDOs.
Different EDOs approximate JI intervals in different ways. In these graphs, errors are given as a percentage of one degree or "step" of that EDO. Thus, the maximum error on the graph is 50%, which would indicate that the prime-factor lies midway between two neighboring EDO degrees.
Error values tend to cluster into groups -- this clustering is plainly visible on the graphs, and I have also divided the lists accordingly, putting a blank row between clusters.
MouseOver the cardinality names of the various EDOs to see a graph of them showing the error from the closest EDO approximation to 11-odd-limit just intonation.