Turtle wins the race? eventually...

I tried reading some books in the library on the Monte Carlo method, trying to find descriptions of simulated annealing. While I didn't understand a lot of it, I did get:
the probability of accepting a worse point is e^(-diff/temperature parameter). that helped.

I have also added in a permutation of the time each round. The code also prints the results each round.

I have also added in the "temperature" that controls acceptance of worse results. Some simulated annealing tutorial said Tnew= alpha * Told was a common annealing schedule, so I decided to adopt that. I would like to research that choice a bit more though. I set alpha as .998 so that it could still move up half way through a 1000 long run, but probably not accept anything near the end.

I decided a 1000 long run based on time. One survey takes about 15 seconds. If I average five surveys per point, then leave about 15 seconds to calculate new point, etc., then 1000 iterations would take 21 hours. I'd like to get it more efficient than that, but I don't know how.

I still need to calculate noise. I also need to restrict time to under 10 hours per semester some how. I think I'll need to keep the time from having two of the same dates some how too.
At least my to-do list is slowly shrinking. I still am trying to read some difficult papers/ books on this subject and experiment optimization.

One paper I'm reading is talking about optimizing future optimizations after 3 or more velocities have been taken. If I could set up that framework now, I'd like to try to implement that into the latter part of the survey if at all possible.


Survey Update
one 9th mag, 2.82 solar mass star in NGC2527 could be added, but SIMBAD doesn't call it a "star in a cluster" like most other stars in clusters