search

jpeterbaker / maxfield

Code for maximizing the Ingress fields on a given set fo portals
GNU General Public License v3.0
165 stars 56 forks source link

Avaible keys optimization #5

Open glupilogin opened 10 years ago

glupilogin commented 10 years ago

Hi, great work with the code there! I had my 1st experience with py but managed to run it (thanks for the detailed how to).

Here is a question: would it be possible to optimize CF/Links configuration for the given keys parameter? Often some portals won't drop as many keys as we like, while others do it everytime. Eventually I end up waiting burnout cylces to collect 1-2 missing keys and having a load of others. So, if I could omit the maxed out CF configuration for the sake of making the best out of the keys I have that would be awesome.

Regards!

jpeterbaker commented 10 years ago

Key proliferation shouldn't be much of a problem anymore since you can't obtain more keys to a portal without dropping or using the old ones. The current code tries to choose a plan that avoids making you harvest more keys than you list in the portal csv file (that's what the "12 tries since improvement" stuff is about). I haven't put any thought into maximizing the number of fields using ONLY the keys that are already available. I think that may have to be an entirely different algorithm.

On Thu, Jan 16, 2014 at 3:01 AM, glupilogin notifications@github.comwrote:

Hi, great work with the code there! I had my 1st experience with py but managed to run it (thanks for the detailed how to).

Here is a question: would it be possible to optimize CF/Links configuration for the given keys parameter? Often some portals won't drop as many keys as we like, while others do it everytime. Eventually I end up waiting burnout cylces to collect 1-2 missing keys and having a load of others. So, if I could omit the maxed out CF configuration for the sake of making the best out of the keys I have that would be awesome.

Regards!

— Reply to this email directly or view it on GitHubhttps://github.com/jpeterbaker/maxfield/issues/5 .

glupilogin commented 10 years ago

Thanks for the reply. What I was aiming for is an optimization algorithm for a situation where you have many friends (say on a organized farm) and everybody drops you keys. Easly I end up with 20 keys for some of the 20 portals I link daily but it happens i have only 1 or none for some others (those lazy droppers). Regards

jpeterbaker commented 10 years ago

I see that if you routinely collect keys from other players, lopsided harvests are likely. Do I understand correctly that you're interested in an algorithm for maximizing AP with a fixed collection of keys?

glupilogin commented 10 years ago

Correct. Im happy with the maxfield way (which equals max AP I guess) BUT for a fixed key structure scenario. Would it be doable? Don't want to take your time too much :) Cheers

jpeterbaker commented 10 years ago

This seems like a difficult problem in general. If you have almost enough keys for maxfield, you pick a maxfield plan that only loses a few fields because of the missing links. If you only have a few keys, you could do a brute force search. I haven't been able to think of a good method for the in-between cases. Maxfield definitely maximizes AP (I have a proof which I haven't gotten around to distributing yet). Let me know if you come up with something. In the meantime, go get those out-of-the-way keys.