[seqfan] A metric and UI for Pandora-like OEIS exploration

[Robert Munafo]

This is a status report.

I have the concordance metric working, and a lot of the utility functions
for the controlling script and user interface. On a 2007 version of OEIS
containing 131553 sequences, a Core 2 Duo laptop can "thumbs-up" a sequence
in 18 seconds, and my 8-core Mac Pro takes less than 4 seconds. (It spawns
as many threads as it can on the hardware, provided that "sysctl -n hw.ncpu"
works.)

The concordance match seems to work at least a little -- given A000213, the
closest match is A000073 (score 0.557). Here are some more pairs of closest
matches. Whether these make sense or not, you be the judge (-:

Primes:
A000040 <-> A000045 (score 0.353) (Fibonacci numbers)

Powers of 2:
A000079 <-> A000110 (score 0.531) (Bell numbers)

Pennies sequence:
A005577 <-> A005575 (score 0.641) (another pennies sequence)

Composite numbers:
A002808 <-> A000720 (score 0.397) (the prime counting function)

The "How four dogs meet in a field" sequence:
A006451 <-> A006454 (score 0.520) (another diophantine equation)

Kaprekar triples:
A006887 <-> A006886 (score 0.512) (the normal Kaprekar sequence)

Greedy Egyptian fractions:
A100140 <-> A000129 (score 0.208) (Pell numbers)

-- 
 Robert Munafo  --  mrob.com