![]() |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Welcome to Deadbeef.com
SearchLinksOpen SourceOtherFavoritesWho's Online?
Misc |
|
Yet Another Double Fanucci Calculator (YADFC) improved06/14/10Yet Another Double Fanucci Calculator (YADFC) improvedI'm embarrassed to admit it, but I have become quite attached to a web based game called Legends of Zork. It is slow, repetitive, and requires little skill. In other words it is just my game playing speed. One of the puzzles in the game is to arrange your Double Fanucci cards to get the most points so that you can win more encounters. I found a very nice Java based calculator called Yet Another Double Fanucci Calculator. It works like a charm, but is it a bit slow, and the genetic algorithm doesn't always find an optimal solution. I have profiled the program, and improved it in several ways.
Download a runnable jar: yadfc-v2.3B-deadbeef.jar I'll keep this page up to date as new versions are released, post any issues to the comments here. Update 6/14/2010
Comments, Pingbacks:
forgive me beeing a noob, but how do you run this program?
Comment from: rajen [Visitor]
Pl. give the download link again. not working.
My problem is that it seems to think that it needs to put stuff into my sidekicks.
Right now my Spirit is 7 with Lamps-0 / Faces-0 / Faces-1 But it's telling me that I should put Bugs-0 / Clocks-0 for a total of 4 in Spirit and putting 3 into sidekicks. Ran it several times and same results Tried the original one, and it's giving me the correct 7...think i'm going with what the original says....
Comment from: Jeremy [Member]
I just ran v2.3B and put in Bugs-0, Lamps-0, Time-0, Faces-0 & 1; 5 cards total. Then I selected Simulator->Run Recursive... and it gave me:
7: Lamp-0, Face-0, Face-1 4: Bug-0, Time-0 Seems right to me.
I dropped the ball on this. If i'd remember to check back I would have been able to send you the exact cards I had. I've gotten several more since, and don't know which ones i had at that time.
All i know is what I said, with what i had at the time (about 20 cards), it was giving me less Spirit than the original calc. I've been running both whenever I've gotten a new card, and the #'s have been consistently the same on both. I'll try experimenting and see if i can replicate it. (my luck, i wont)
Ok, tried something...on the original I set up several cards
0 bug 0 lamp 0 maze 0 hive 0-3 inkblot 0 ear 0-1 time 0 scythe 0 zurf 0 plunger 0-1 face ran it 5 times. and Sidekick would change in the "Best solution" 1st time 0 scythe for a 2 2nd time 0 Lamp for a 2 3rd time 0 lamp AND 0 scythe for a *3* 4th time 0 Maze for a 2 5th time 0 lamp AND 0 scythe for a *3* again and i just kept refreshing and kept happening... Interestingly the Mind/Body/Spirit/Gambit values never changed. (8/8/4/28) So, the original is definitely flawed. Ran the same card combo on your version and always got the same result. Now...if only i could figure out what i DID have that was giving me less spirit in yours :( Oh well...guess it don't really matter.
Using your version
Copy and pasted the cards from Zork, so it don't mean I have the 6 face card, i have 0-5 for a total of 6....etc. You currently own 30 Fanucci cards Suit-faces 6 Suit-bugs 1 Suit-lamps 1 Suit-mazes 3 Suit-fromps 1 Suit-hives 1 Suit-inkblots 5 Suit-ears 2 Suit-time 4 Suit-scythes 1 Suit-zurfs 1 Suit-books 1 Suit-plungers 2 Suit-tops 1 Run using Genetic This is with the standard options that comes with it. Run 1: M-48/B-28/S-16/SK-10/FG-50 Run 2: M-48/B-28/S-16/SK-9/FG-50 Run 3: M-48/B-28/S-16/SK-9/FG-50 Run 4: M-48/B-28/S-16/SK-10/FG-50 Run 5: M-48/B-28/S-16/SK-10/FG-50 Mind, Body, Spirit, and Gambit never change #'s (tho the cards DO change a few times) and Sidekick floats from 9 to 10 Decided to up the ante and take a chance, so upped everything in the options because i have NO idea what they are and there's no manual :) Pop Size 4096 Elitism Rate 100% Mutation rate 100% Iterations 512 Sets 5 Maxima 64 Run 1: M-46/B-28/S-18/SK-10/FG-49 Run 2: M-46/B-30/S-17/SK-10/FG-47 Run 3: M-47/B-28/S-16/SK-14/FG-48 Run 4: M-40/B-28/S-25/SK-10/FG-50 Run 5: M-39/B-29/S-27/SK-10/FG-50 Everything changed at least once I also ran these tests with Recursive and got the various ones as well. Honestly, don't know the difference between the 2...but don't seem to be any
Comment from: Jeremy [Member]
The Genetic algorithm has some randomness in there. It picks a random starting point and hill-climbs to a solution. The Recursive algorithm, on the other hand, considers all cases and should come up with the same solution every time. Once the recursive one has proven to be stable and fast, I plan to delete the other.
Comment from: Aquila76 [Visitor] · http://legendsofzork.com/heroes/359624-aquila76
This is great! Thanks for keeping this DFC project alive. Working great for me (68 cards).
I'm sorry, but I'm confused here. When I found this site, I had already found and downloaded the YADFC program. I understand your intentions, but do not do any programming or development. When I downloaded your "runnable jar" the compressed file simply had three folders named "META-INF" "com" and "net" looking at the download there were no instructions included and it appeared if I looked at the source code, it did mot make it clear how this was set up, whether something needed to be compiled manually, if running the original JAW would do anything with the files, where those downloaded files should be (should they be placed with the original program in their folders or extracted differently) etc?
While I do have technical skills, program compilation and development are not two of them. If you'd gladly guide me and the other user Joe the visitor, could you advise us how we are to set up the program once we downloaded the compressed file. Then since I am familiar more with statistics and algorithms, perhaps my feedback will be more productive next post.
D'OH! My download was tanslating it as a compressed .zip and manually downloading it as a jar fixed that. Wiping egg off my face and noticing no edit on the post site :)
Leave a comment:
|
|
Sponsored LinksTop Articles
Categories
Archives
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||