kennethshackleton/skpokereval

7-card Texas Hold'em hand evaluator

skpokereval's Issues

dll with exported functions

Not so much an issue as a request..

Is there any way to get your code compiled into a DLL with the functions exported? - I've written a GUI in Clarion :) which shuffles and deals cards for Texas Hold'em (for my own use of course). I would like to be able to simply call your function to compare the cards for two players. 2-pocket cards each and the 5 board cards - so 7 cards per player. A value is then returned and the winner is the one with the highest/lowest value...and then possibly also be able to get the hand type like flush, straight, etc.. Is that do-able at all?

thanks!
W

is the old 5 card eval code unuseable?

It would be useful to use for plo ranking since 7 card eval doesnt enforce 2 holecards playing.

6 card eval

Hi, my tests show this evaluator is blisteringly fast. Phenomenal work.

Any chance you can add a 6 card version?

Is this a +- fastest Texas Holdem evaluator?

Hi can you tell how perfomamnce of this evaluator is compared to other known and available is faster, lower +- the same? Thanks.

Tests Slow

Some of the tests run slow on my machine, is this normal?

Running main() from gtest_main.cc
[==========] Running 19 tests from 2 test cases.
[----------] Global test environment set-up.
[----------] 18 tests from FiveEvalTest
[ RUN ] FiveEvalTest.HighCard
[ OK ] FiveEvalTest.HighCard (1 ms)
[ RUN ] FiveEvalTest.WorstPairBeatsBestHighCard
[ OK ] FiveEvalTest.WorstPairBeatsBestHighCard (0 ms)
[ RUN ] FiveEvalTest.Pair
[ OK ] FiveEvalTest.Pair (0 ms)
[ RUN ] FiveEvalTest.WorstTwoPairBeatsBestPair
[ OK ] FiveEvalTest.WorstTwoPairBeatsBestPair (0 ms)
[ RUN ] FiveEvalTest.TwoPair
[ OK ] FiveEvalTest.TwoPair (0 ms)
[ RUN ] FiveEvalTest.WorstTripleBeatsBestTwoPair
[ OK ] FiveEvalTest.WorstTripleBeatsBestTwoPair (0 ms)
[ RUN ] FiveEvalTest.Triple
[ OK ] FiveEvalTest.Triple (0 ms)
[ RUN ] FiveEvalTest.WorstStraightBeatsBestTriple
[ OK ] FiveEvalTest.WorstStraightBeatsBestTriple (0 ms)
[ RUN ] FiveEvalTest.Straight
[ OK ] FiveEvalTest.Straight (0 ms)
[ RUN ] FiveEvalTest.WorstFlushBeatsBestStraight
[ OK ] FiveEvalTest.WorstFlushBeatsBestStraight (0 ms)
[ RUN ] FiveEvalTest.Flush
[ OK ] FiveEvalTest.Flush (0 ms)
[ RUN ] FiveEvalTest.WorstFullHouseBeatsBestFlush
[ OK ] FiveEvalTest.WorstFullHouseBeatsBestFlush (0 ms)
[ RUN ] FiveEvalTest.FullHouse
[ OK ] FiveEvalTest.FullHouse (1 ms)
[ RUN ] FiveEvalTest.WorstQuadBeatsBestFullHouse
[ OK ] FiveEvalTest.WorstQuadBeatsBestFullHouse (0 ms)
[ RUN ] FiveEvalTest.Quad
[ OK ] FiveEvalTest.Quad (0 ms)
[ RUN ] FiveEvalTest.WorstStraightFlushBeatsBestQuad
[ OK ] FiveEvalTest.WorstStraightFlushBeatsBestQuad (0 ms)
[ RUN ] FiveEvalTest.StraightFlush
[ OK ] FiveEvalTest.StraightFlush (0 ms)
[ RUN ] FiveEvalTest.SevenCardHand
[ OK ] FiveEvalTest.SevenCardHand (116314 ms)
[----------] 18 tests from FiveEvalTest (116316 ms total)

[----------] 1 test from SevenEvalTest
[ RUN ] SevenEvalTest.CompareWithFiveEval
[ OK ] SevenEvalTest.CompareWithFiveEval (72018 ms)
[----------] 1 test from SevenEvalTest (72018 ms total)

[----------] Global test environment tear-down
[==========] 19 tests from 2 test cases ran. (188334 ms total)
[ PASSED ] 19 tests.

This is not an issue

Sorry for abusing your issues list but I don't know where to write.

I'd like to just thank you for your amazing tool.

I wrote something similar (but based on bitboards) and SKPokerEval has been really useful to fix a couple of subtle bugs.

Now that my scoring comparison matches with SKPokerEval , I feel confident I have fully validated my little tool. Thanks again and keep up the good work!

FiveEval large and invalid ranks

I see that FiveEval was moved from the main repo into tests and that would appear that you no longer support it but I have a use case were I would like to evaluate the strength of 5 card poker hands. I have noticed that when evaluating a large number of hands in a row sometimes FiveEval will start returning about 50% of it's results as large invalid scores. Do you have any ideas as to what could be causing this? I have verified that the hands I am passing to FiveEval are valid 5 card poker hands(no duplicates)

It also perform an evaluation of a 5 cards hand?

If yes, how can i get it?
Thanks

Should include rank ranges in the readme

Gaps in the enums introduce entropy to deep learning systems.
I found this useful while making onehots
These are sparse ranges
0 == error
High Card: 49 - 1277
Pair: 1296 - 4137
Two Pair: 4141 - 4995
Three of a Kind: 5004 - 5853
Straight: 5854 - 5863
Flush: 5864 - 7140 -- flushes seem to be uniform across suits, two equal flushes will score the same
Full House: 7141 - 7296
Four of a Kind: 7297 - 7452
Straight Flush: 7453 - 7461
Royal Flush: 7462 - 7462
//same as straight flush

Total distinct ranks possible:
High Cards: 407
Pair: 1470
Two Pair: 763
Three of a Kind: 575
Straight: 10
Flush: 1277
Full House: 156
Four of a Kind: 156
Straight Flush: 10