Why not implement the standard harmonic mean distance calculation as paper said? about aflgo HOT 1 CLOSED

In Figure4 the harmonic mean is calculated as 1/(1/2 + 1/2)， but it should be like:
n/( 1/2+1/2 + ... + 1/2 )

The implementation in code is right （the "n" in numerator）.

Something else in code which is a bit different from the paper, "+1" was added to denominator, maybe is to avoid the div 0.

And the final basic blocks's distance is the average of each situation, instead of considering in different situations as Formula 2 in paper. I guess this is to take more situations into consideration, instead of merely the shortest distance.

The line you pointed is to calculate the mininum distance, as there may be several nodes returned from find_nodes().(not sure)

from aflgo.