[Resolved] 50 Counties with a Top10 favorite cache found challenge

Hi

I think I have an idea for a new class of challenges. Today most favorite challenges are simply the top10 in the state or some number of finds with something like that. Is it possible to to make a checker to find if the cacher has found a top 10 favorited cache in 50 different counties. The purpose is to encourage finding quality caches spread out across the country or world. Project-GC has a really cool statistic-Top Favorites menu item that you can then add filters to select top caches in states or further down to counties. If this could be automated it would be pretty cool in my opinion and encourage to find popular caches all over the place.

Thanks

It would be possible to write a script for this.

Before that happens, though, you should probably run this by a reviewer. It may be considered to contravene Guideline 4:

Quote

Challenge cache criteria

• must come from information broadly available on Geocaching.com such as on the statistics page, cache placement dates, types, attributes, souvenirs, etc.
• must be verifiable through information on Geocaching.com.

Edited 1 time(s). Last edit at 06/01/2018 06:42AM by sumbloke. (view changes)

I have a script which has been originally made for the most favorited coutries but it works also with counties.

https://project-gc.com/Challenges//35155

This allows currently only the most favorited cache but it could be modified for more than one if the challenge is allowed by the reviewer.

There is one potential problem. The tool takes quite long to find qualifying caches and it is possible that with huge amount of finds it will not give any results.

Thanks for the quick responses and letting me know it was possible. Will see if they will publish it next. Just getting bored with the existing challenges and was trying to come up with something novel and encourage getting good caches all over the place. The top caches tend to be tourist draws and clumped here in Oregon around the original stash & GC12/GC16/GC17. Or in Washington around GCHQ. Wanted to encourage getting the truly great caches out there and not just the big ones.

Hi there just got word from the reviewer that they will probably publish it. He wanted to see the results and ensure it handles archived caches which project-gc statistics page displays nicely.

The checker you included above is fabulous and looks like it is very close to the desired challenge. BTW I ran on T&A137 and if gave an exception euro.

Can you please proceed.

Do you already have the GC-code for the cache? The checker needs it.

Yes GC7R26B Top10 favorite in 50 counties challenge

I have created two checkers to test. One with full output and another with brief output. The brief versio is faster and may not time-out when full version does. If the challenge could be limited to specific country or states it will be faster. The problem is that getting the most favorited caches from every visited county will take some time and there is practically nothing we can do for making it faster.

Full version: https://project-gc.com/Challenges/GC7R26B/35399
Brief version: https://project-gc.com/Challenges/GC7R26B/35398

Please test and let us know whether they are suitable for you.

There are some tricks you can play to speed this up.

The first, and the one you probably know, is to just stop when you have enough caches/counties. Just show they have qualified and be done with it.

Another is to have the methods extended. PGC.GetFavorites doesn't support an array of counties. This is a mod that can be requested and likely will go through. (For GetOldest, I think I made this request) Also it would be a good time to check that GetFavorites is using the columnar database in the backend which is much faster. The you could request multiple counties at a time and filter the list. Some prior conversation on this is here.

Remaining, there are two games you could play. (Assuming that the county mod goes through) The first is to request 1000 favorites from all the counties, and then parse that list to get what top 10's you can. Then over more passes, reduce the number of counties in the list as you find caches and have counties you have collected the ten favorites. You might care to hash things by region as well. Grab 1000 from a region and then do the checks to see if you have enough info.

Magnus has been receptive in the past to reasonable mods. Given that you have a working test case, he should be quite understanding.

Hope this helps.

arisoft Wrote:
-------------------------------------------------------
> I have created two checkers to test. One with full
> output and another with brief output. The brief
> versio is faster and may not time-out when full
> version does. If the challenge could be limited to
> specific country or states it will be faster. The
> problem is that getting the most favorited caches
> from every visited county will take some time and
> there is practically nothing we can do for making
> it faster.
>
> Full version:
> https://project-gc.com/Challenges/GC7R26B/35399
> Brief version:
> https://project-gc.com/Challenges/GC7R26B/35398
>
> Please test and let us know whether they are
> suitable for you.

> There are some tricks you can play to speed this up.

The brief version is currently fast enough for Alamogul. I think this is not a critical problem for the challenge itself.

It could also possible to limit the region smaller and make a precalculated table for estimations per county to select the most potential ones to be processed first.

I agree. I think we cross posted while you were optimizing. nicely done.

arisoft Wrote:
-------------------------------------------------------
> > There are some tricks you can play to speed this
> up.
>
> The brief version is currently fast enough for
> Alamogul. I think this is not a critical problem
> for the challenge itself.
>
> It could also possible to limit the region smaller
> and make a precalculated table for estimations per
> county to select the most potential ones to be
> processed first.

I want to first say thanks they were both awesome! I definitely appreciate this and understand a small part the challenges and the run time. It was great seeing the output of the full version. Was able to visually see what it was doing and also ensure it was selecting the archived caches which the reviewer was interested in. I did see it crash on user T&A137 if you could take a took at. I'd say that the brief version is really all that is required even though its cool to see the other locations and caches but I doubt everyone will use the checker as a planning tool. The project-gc statistics page is probably better to do that there. Since we are going wth the brief version would it be feasible to include the favorite# rather along with the number of favorites.


Thanks again this is great.

So playing around a little more. Looks like the archived caches are excluded. Was looking a user SpockProf in the full version. The reviewer sounds like they really want archived caches included.

I see:
Tillamook County (OR) (♥342): GCA5 Hembre Ridge (2014-10-09)
Tillamook County (OR) (♥232): GC1EWZ9 Tunnel of Fun (2014-03-24)
Tillamook County (OR) (♥116): GC2A68 Visit Aunt Helen (2014-01-02)
Tillamook County (OR) (♥74): GC3M4NX Picea Sitchensis Octopoda (2014-01-02)
Tillamook County (OR) (♥72): GC142 The Burn (2014-08-09)
Tillamook County (OR) (♥54): GC1KP5Q Great Grandma Tree (2017-05-29)
Tillamook County (OR) (♥49): GC2C4DQ Cascadia Challenge (2014-10-10)
Tillamook County (OR) (♥47): GC3QTDC Indiana Jones and the Tunnel of Doom (2014-01-02)
Tillamook County (OR) (♥45): GC1T38P Ocean Ridge (NOT FOUND)
Tillamook County (OR) (♥44): GC2PHJ9 It's a Wonderful Life (2015-12-31)

But looking at the statistics page for Tillamook. #7 & #11 are archived and It's a wonderful Life is 12th. Personally I like active caches only but understand the desire to include archived. I was not going to be the cacher police monitor that in this case #12 would be acceptable as the spirit of the challenge is met. So maybe the happy medium is to allow both. What do you think?

The way the oldest challenges have been implemented is to include the specified number of active caches and count any archived cache older than them. Maybe something like that would work for this challenge?

In other words, have the tenth-highest FP cache at the time provide the FP threshold for that county and any find in that county with at least that many FP would satisfy the criteria. In your example for Tillamook County, any cache with at least 44 FP would count, including #11 in the list that includes archived caches. It would also help for counties which might have a tie for tenth.

I completly forget the archive status. It makes things complicated as there was no rule how to handle them. It is not right to require to find any archived cache but allowing them as on option is another problem because we need to decide whether a cache is in the top ten or not. This need rewriting the rules to the challenge as sumbloke suggested.

I have to say that I do not fully understand the logic how PGC finds the most favorited caches but for some reason the result seems to be much better now.

There is one important thing to notice that caches with the same number of favorites shares the same rank. The checker do not try to limit the number of caches to the limit but the number of ranks instead. This may need to be fixed somehow if it is important.

Tillamook County (OR) (#1♥342): GCA5 Hembre Ridge (2014-10-09)
Tillamook County (OR) (#2♥232): GC1EWZ9 Tunnel of Fun (2014-03-24)
Tillamook County (OR) (#3♥116): GC2A68 Visit Aunt Helen (2014-01-02)
Tillamook County (OR) (#4♥74): GC3M4NX Picea Sitchensis Octopoda (2014-01-02)
Tillamook County (OR) (#5♥72): GC142 The Burn (2014-08-09)
Tillamook County (OR) (#6♥54): GC1KP5Q Great Grandma Tree (2017-05-29)
Tillamook County (OR) (Archived #7♥53): GC28YMY Up It Comes (NOT FOUND)
Tillamook County (OR) (#8♥49): GC2C4DQ Cascadia Challenge (2014-10-10)
Tillamook County (OR) (#9♥47): GC3QTDC Indiana Jones and the Tunnel of Doom (2014-01-02)
Tillamook County (OR) (#10♥45): GC1T38P Ocean Ridge (NOT FOUND)
Tillamook County (OR) (Archived #10♥45): GC1P8HJ War Bunker (2014-01-02)
Tillamook County (OR) (#11♥44): GC2PHJ9 It's a Wonderful Life (2015-12-31)

I don't know how or why the checker adds the last line automatically to the correct position. For some reason it does show only archived caches found by the user without any instruction to do so ::o the previous version of the checker just didn't allow them to be used for qualifying.

Testing the checker is a little bit difficult because there is not so many archived and top favorited caches. I hope that you can find suitable test objects to verify it thoroughly.



Edited 3 time(s). Last edit at 06/12/2018 11:29AM by arisoft. (view changes)

After some more testing I now understand how the most favorited list is working. It adds archived caches to the list but do not count them for the limit. That is why there is more than ten ranks. For example Tillamook have #11 because there is one archived cache at position #7 which is not counted for the limit. Limit is number of caches unless there is the same number of favorites at the last rank.



Edited 1 time(s). Last edit at 06/12/2018 10:17AM by arisoft. (view changes)

The problem with "T&A137" was some finds without any county name. Fixing this did not help because there is also too many counties in the full listing but the brief version works properly because it stops finding more counties when the requirement is met. I think this problem is solved.

GC15RQB is a top favorited and archived cache.

The latest version this checker looks great! Really appreciate the work on this one. Is this ready to go?


Thanks a lot hopefully you enjoyed the challenge as well.

I think it is ready for use. You are going to use the brief version so I can remove the full one?

Yes the brief version will be used. The full version fails for folks that have traveled extensively.

Thanks again

Cache is ready to go, reviewer has approved the challenge now we just need the checker to be released. Can I get the numbers reset?

> Brief version:
> https://project-gc.com/Challenges/GC7R26B/35398

Huh

If I search for GC7R26B in the challenge checker list it does not come up. Wish to use the brief version that was supplied above. Can the stats be reset instead of having my testing runs so it says zero pass and zero fail.

Not that I am aware of.

No problem. Thanks for all the work everyone has done on this. The brief version of the checker is the one we will be using and I have included a link to it in my cache page. I think from my side this is completed.

Are you only able to run the checker one time? I ran it once, and found I qualified. I left the page before copying it down, and when I try running it a second time, it times out with an error but no information about what kind/type of error.

I have seen sometimes that a checker may crash without explanation more than once but later it works again. This checker is very database intensive and the behaviour may relate to the current load level of the database server.

I tried this checker now three times without success but I guess that it will work later if I try again.

Works great consider this done! Thanks again!