Yes, I know.

The numbers you proposed would have made it worse.

Currently the program looks for anything within +- 3 windows in time, up to 10 times the radius in space and any magnitude at all. The search radius is increased in 10% increments for probability calculations since one does not predict exact values for things.

But I get too many hits to compute in a reasonable time (in some cases) so I need a weed-out routine. I can't do it without a major rewrite because right now it remembers everything and does probability last and I run out of memory at about 400 quakes.

Always something....

Roger