×

To be able to write in the forum you need to authenticate. Meanwhile it's read-only.

Re: GC42CNY 12-12-12: A Challenge? (Rotorua)

[Resolved] GC42CNY 12-12-12: A Challenge? (Rotorua)
August 14, 2016 12:07PM
Would appreciate if someone would show mercy and write a checker for this Challenge
cache

//Rickard
Re: GC42CNY 12-12-12: A Challenge? (Rotorua)
August 14, 2016 07:09PM
The challenge requires a special script and I doubt if this will be made in the short terms.
This is likely to be low on most script-writer's priorities since the script would have limited usefulness.
Re: GC42CNY 12-12-12: A Challenge? (Rotorua)
August 15, 2016 08:41AM
And it cant be done with all rules
There is no way for a script to know when a cache has been disabled. And can only now last archive time
Quote

The logged Caches must all be proven to be active and available at the date of logging this puzzle.
e.g: You cannot include a 'find' log for a cache that is unavailable, archived, needs maintenance, or simply has a string of DNF's. Ditto, you cannot include a logged 'attend' on an archived event (The event cannot have been held more than 12 days prior to logging this cache).

If written the script might be to slow since you have to test all combinations of 12 caches find in a 12 day windows. With many find it will be slow.
The script cant be reused because it would not be allowed today
Re: GC42CNY 12-12-12: A Challenge? (Rotorua)
August 22, 2016 07:15AM
As cache CO, here's my input...

Most of the rules could (& will be) be discarded - when the challenge was published, there was a problem with a few - 1 in particular - cachers who specialised in pre-dating caches that had been archived or were disabled / DNF, thus claiming all sorts of things. Eventually, gc.com closed them down. (He had multiple profiles, it transpired, and an entire country annoyed with him. I was probably grumpy at the time about him so made sure he was specifically excluded, using his 'fake' finds & fake caches that he published, FTF'd under one or more of his alt.profiles, then no one else could find them.) After all, it is a D5/T5 cache and was set up for a special date 12-12-12.

However, the challenge requirements in terms of a checker are:
  1. Find 12 caches in 12 days
  2. Must be at least 2 unique types
  3. Total D must equal Total T
  4. Total D must be 24 or 36 or 48 or 60

First 2 requirements are easy.
The 3rd & 4th are a question for the experts here to decide upon.

If not feasible, I'd kindly ask that this thread be closed.

Cheers, GC.

Edit: Cache page rewritten to meet pre-2016 challenge rules.



Edited 1 time(s). Last edit at 08/22/2016 09:46AM by GenCuster. (view changes)
Re: GC42CNY 12-12-12: A Challenge? (Rotorua)
July 17, 2017 07:06PM
I felt like a bit of a challenge, so...

As mentioned, (a) and (b) are easy. (d) is easy too. The problem is (c), since it becomes an exponential problem once you've found more than 12 caches in those 12 days, since you have to look through all the possible permutations. For example, if you found 20 caches, you have 8*7*6*5*4*3*2 = 40320 permutations to check. Now if you have a few of those, you run out of the allowed checker CPU time very quickly. However, I managed to construct a checker by first searching through the shorter options and hoping for the best before it's forced to delve into the longer options. This means this checker will find ONE match at a random date in your geocaching history. Yes, I first tried looking for the first or last match, but that didn't work (too much CPU usage). That said and done, you may try this custom built checker:

http://project-gc.com/Challenges/GC42CNY/27654

Please let me know what you think?
Re: GC42CNY 12-12-12: A Challenge? (Rotorua)
July 17, 2017 10:57PM
It gives me a list of just traditionals...
Re: GC42CNY 12-12-12: A Challenge? (Rotorua)
July 17, 2017 11:02PM
It does check the types. One of the optimisations to limit the search grid is to skip anything past 12 of the same DT for a group of 12 days and that's where your 2nd type might appear to fall off the list. I have pondered that this might happen and you've now proven it. Will have to ponder how to fix that. Maybe I simply need to insert the 2nd type as the first one for the 12 day period so it gets selected first. Let me try that.

Nope, it seems my type test was not used. Try it again now?
Hmm, not working. Debugging...
And so we learn: 0/1 ~= false/true. Try it again now, please?



Edited 3 time(s). Last edit at 07/17/2017 11:15PM by pieterix. (view changes)
Re: GC42CNY 12-12-12: A Challenge? (Rotorua)
September 10, 2016 05:43AM
maybe it would be helpful to create a checker only for the first two tasks? It will reduce significantly the possibilities. The last two requirements can do manually.
Re: GC42CNY 12-12-12: A Challenge? (Rotorua)
July 27, 2017 07:16AM
Any feedback on the checker as mentioned in this thread?

http://project-gc.com/Challenges/GC42CNY/27654
Re: GC42CNY 12-12-12: A Challenge? (Rotorua)
August 11, 2017 08:47AM
Many thanks. It's really appreciated.

I placed an Owner Maintenance Log on my cache with these details so potential finders can have a test of this.

Worked OK for me and other random qualifiers that I selected.

Error if a cacher does not currently qualify - see Aiden29.

Also some requests please:
1. Could the output be formatted so it can be cut and pasted into a log? Currently the formatting does not lend itself to that with the gifs. Perhaps cache type as text, and at the end? e.g.:

D / T / Date Found / GCCODE / Cache Name / Type
3.5 1.5 2013-08-23 GC3JNDY A clue to a bridge Mystery
2.5 2.5 2013-08-24 GC3NDH6 Ammoudi Traditional
etc
24 24

Currently the cut and pasted log would be:
# date t D T gc name
1 2013-08-23 8.gif 3.5 1.5 GC3JNDY A clue to a bridge
2 2013-08-24 2.gif 2.5 2.5 GC3NDH6 Ammoudi
etc
24 24


I'm not worried about having links in the output, because unless it's a direct link to the cacher's log, it doesn't really achieve anything, as the Geochecker has already 'proved' it. Awesome!

2. Possibility of working backwards by date? As an option??? So latest qualification would be logged. Reason: Most cachers log intention then start to find qualifying caches. Some cachers just use whatever qualifying caches they have. Cacher DorisBear is an example - Geochecker says qualified in 2013, but they actually used caches in March 2017.

3. Possibility of starting from a specific date or cache ID? Again, so the cacher can use specific caches as a starting point. But more importantly, as the Geochecker is now, it finds 1st qualifying 12 caches and stops. Which may result in a D 24 / T 24 log. But if the cacher has a specific D 36 / T 36 set of caches (or even a D 48 / T 48), it would be amazing if they could select a starting point to ensure those are specified. Cacher chudles is an example - Geochecker qualified in 2013 as D 24 / T 24, but cache logged with caches found in April 2015, as D 48 / T 48.

I think that requests 2 & 3 may also speed up the searching. If not possible to add an option to select start point, I would (respectfully) ask if start point could be current date and work backwards, rather than start from the 1st cache a particular cacher has found.

And Thanks again,
GenCuster
Re: GC42CNY 12-12-12: A Challenge? (Rotorua)
August 11, 2017 09:36AM
Preface: will have to study my code and see if there's a way to more quickly find a solution. This script was written in my early days as script writer and I've learnt a few things since then.

1. I've added a text formatted logging option and left the HTML version as is for now. Give it a spin and let me know?

2. That's one of the options I tried to speed up the searching for solutions, but I didn't have any luck back then. However, see preface above. Not sure what the implications would be if the script is linked multiple times to the same challenge, ie. to give you different versions of the same script, eg. one that searches from the start and one that searches from the end. Any comment from Target, sumbloke, vogelbird?

3. It's not technically possible to get user supplied parameters to the script, so there's no way to implement this. Could possibly allow the script to continue running once it's found the first match to attempt to find other higher DT sums, but I can't promise this will work.

Will give further feedback once I've had time to take the script apart and try a few things.
Re: GC42CNY 12-12-12: A Challenge? (Rotorua)
August 11, 2017 11:38AM
After reviewing my code, a note about 2: it seems I'm not searching by date, but actually by number of finds per 12 day period, hoping to a solution when there's less permutations to examine, before heading into deeper water and look for solutions when there's more finds per day. (More finds per 12 day period = more permutations to test). Will however review my code and see if there's a better way to do the searching.
Re: GC42CNY 12-12-12: A Challenge? (Rotorua)
August 11, 2017 02:22PM
I've speeded up the script a bit and modified it to continue hunting for bigger scores until it gets to 25 sec (30 sec is the maximum runtime for a script) at which time it will abort and show the highest score. This is not a guarantee that it will find the biggest score, since it always attempts to find the quickest solution for the set of finds for each 12 day period with 12 or more caches. Which means it will always find A solution, but not necessarily THE solution. Tried to push the issue by sorting the higher DT caches to the front of the pile to be checked, but I'm not sure if that would be enough of a push. Given the run time limitation there isn't much more I can do about it, unless I dream of another hunting method. Will ponder this over the weekend.

PS: chudles now gets to 36 and Aiden29 gets a negative instead of a runtime error.
Sorry, only registered users may post in this forum.

Click here to login