Sebastiano Vigna

If there is interest, I might commit harmonic eccentricity, too, which follows the same idea (harmonic mean of distances from x to all other nodes).

Actually, no, I can't, it's a fairly obvious "harmonic" reworking of eccentricity. You can use harmonic means in closeness, and you get harmonic centrality, in diameter computation, and you get...

Yeah, I was dubious whether to add a note. I did now. In the paper the harmonic diameter is called "connectivity length". But the name "harmonic diameter" is standard in...

Yes. So, you could define the standard diameter as the maximum distance between k points for every choice of k points and let k go to infinity, as they do...

Incidentally, that function is really weird. It states a formula but does not compute that formula. If I apply it to a directed graph with a single arc 0→1, it...

Sorry—is there anything else I have to do? Or is this going to be included in NetworkX at some point?

OK, so now this is conflicting. Should I make the effort of solving the (trivial) conflicts or are you guys not interested in this new code?

Incidentally, I realized I'm using the term probably because I found it in this 2003 paper by Fogaras: Fogaras, D. (2003). Where to Start Browsing the Web?. In: Böhme, T.,...

This is another thing to maintain and update... what's the advantage?

Done in 85bb33b1bfde34dac21005037951b615268f0083. Let me know if it works for you...