Py-Hanabi/sat.py

341 lines
16 KiB
Python
Raw Normal View History

2023-02-28 14:19:01 +01:00
from pysmt.shortcuts import Symbol, Bool, Not, Implies, Iff, And, Or, AtMostOne, ExactlyOne, get_model, get_atoms, get_formula_size, get_unsat_core
from pysmt.rewritings import conjunctive_partition
import json
from typing import List
from compress import DeckCard, Action, ActionType, link
STANDARD_HAND_SIZE = {2: 5, 3: 5, 4: 4, 5: 4, 6: 3}
NUM_STRIKES_TO_LOSE = 3
# literals to model game as sat instance to check for feasibility
# variants 'throw it in a hole not handled', 'clue starved' and 'up or down' currently not handled
class Literals():
# num_suits is total number of suits, i.e. also counts the dark suits
# default distribution among all suits is assumed
def __init__(self, num_players, num_suits, num_dark_suits=0):
assert ( 2 <= num_players <= 6 )
## some game parameters
self.num_players = num_players
self.num_suits = num_suits
self.num_dark_suits = num_dark_suits
self.hand_size = STANDARD_HAND_SIZE[num_players]
self.num_strikes = NUM_STRIKES_TO_LOSE
self.deck_size = 10 * num_suits - 5 * num_dark_suits
self.distributed_cards = self.num_players * self.hand_size
self.draw_pile_size = self.deck_size - self.distributed_cards
## maximum number of moves in any game that can achieve max score
# each suit gives 15 moves, as we can play and discard 5 cards each and give 5 clues. dark suits only give 5 moves, since no discards are added
# number of cards that remain in players hands after end of game. they cost 2 turns each, since we cannot discard them and also have one clue less
# 8 clues at beginning, one further clue for each suit but one (the clue of the last 5 is never useful since it is gained in the extra-round)
# subtract a further move for a second 5-clue that can't be used in 5 or 6-player games, since the extraround starts too soon
self.max_moves = 15 * num_suits - 10 * num_dark_suits \
- 2 * num_players * (self.hand_size - 1) \
+ 8 + (num_suits - 1) \
+ (-1 if num_players >= 5 else 0)
###
# note that we generate 'literals' always one out of boundary and set them to explicit truth values. This makes sat formulation easier but has no actual overhead in solving it
# move are numbered starting with 0
# clues[m][i] == "after move m we have at least i clues"
self.clues = {
-1: { i: Bool(i < 9) for i in range(0, 10) } # we have 8 clues after turn -1
, **{
m: {
0: Bool(True), # always at least 0 clues
**{ i: Symbol('m{}c{}'.format(m, i)) for i in range(1, 9) },
9: Bool(False) # never 9 or more clues. This will implicitly forbid discarding at 8 clues later
}
for m in range(self.max_moves)
}
}
# strikes[m][i] == "after move m we have at least i strikes"
self.strikes = {
-1: {i: Bool(i == 0) for i in range(0, self.num_strikes + 1)} # no strikes when we start
, **{
m: {
0: Bool(True),
**{ s: Symbol('m{}s{}'.format(m,s)) for s in range(1, self.num_strikes) },
self.num_strikes: Bool(False) # never so many clues that we lose. Implicitly forbids striking out
}
for m in range(self.max_moves)
}
}
# extraturn[m] = "turn m is a move part of the extra round or a dummy turn"
self.extraround = {
-1: Bool(False)
, **{
m: Bool(False) if m < self.draw_pile_size else Symbol('m{}e'.format(m)) # it takes at least as many turns as cards in the draw pile to start the extra round
for m in range(0, self.max_moves)
}
}
# dummyturn[m] = "turn m is a dummy nurn and not actually part of the game"
self.dummyturn = {
-1: Bool(False)
, **{
m: Bool(False) if m < self.draw_pile_size + self.num_players else Symbol('m{}dt'.format(m))
for m in range(0, self.max_moves)
}
}
# draw[m][i] == "at move m we play/discard deck[i]"
self.discard = {
m: {i: Symbol('m{}-{}'.format(m, i)) for i in range(self.deck_size)}
for m in range(self.max_moves)
}
# draw[m][i] == "at move m we draw deck card i"
self.draw = {
-1: { i: Bool(i == self.distributed_cards - 1) for i in range(self.distributed_cards - 1, self.deck_size) }
, **{
m: {
self.distributed_cards - 1: Bool(False),
**{i: Symbol('m{}+{}'.format(m, i)) for i in range(self.distributed_cards, self.deck_size)}
}
for m in range(self.max_moves)
}
}
# strike[m] = "at move m we get a strike"
self.strike = {
-1: Bool(False)
, **{
m: Symbol('m{}s+'.format(m))
for m in range(self.max_moves)
}
}
# progress[m][card = (suitIndex, rank)] == "after move m we have played in suitIndex up to rank"
self.progress = {
-1: {(s, r): Bool(r == 0) for s in range(0, self.num_suits) for r in range(0, 6)} # at start, have only played rank zero
, **{
m: {
**{(s, 0): Bool(True) for s in range(0, self.num_suits)},
**{(s, r): Symbol('m{}:{}{}'.format(m, s, r)) for s in range(0, self.num_suits) for r in range(1, 6)}
}
for m in range(self.max_moves)
}
}
## Utility variables
# discard_any[m] == "at move m we play/discard a card"
self.discard_any = { m: Symbol('m{}d'.format(m)) for m in range(self.max_moves) }
# draw_any[m] == "at move m we draw a card"
self.draw_any = {m: Symbol('m{}D'.format(m)) for m in range(self.max_moves)}
# play[m] == "at move m we play a card"
self.play = {m: Symbol('m{}p'.format(m)) for m in range(self.max_moves)}
# play5[m] == "at move m we play a 5"
self.play5 = {m: Symbol('m{}p5'.format(m)) for m in range(self.max_moves)}
# incr_clues[m] == "at move m we obtain a clue"
self.incr_clues = {m: Symbol('m{}c+'.format(m)) for m in range(self.max_moves)}
def solve(deck: List[DeckCard], num_players=5):
num_suits = max(map(lambda card: card.suitIndex, deck)) + 1
num_dark_suits = (len(deck) - 10 * num_suits) // (-5)
ls = Literals(num_players, num_suits, num_dark_suits)
valid_move = lambda m: And(
# in dummy turns, nothing can be discarded
Implies(ls.dummyturn[m], Not(ls.discard_any[m])),
# definition of discard_any
Iff(ls.discard_any[m], Or(ls.discard[m][i] for i in range(ls.deck_size))),
# definition of draw_any
Iff(ls.draw_any[m], Or(ls.draw[m][i] for i in range(ls.distributed_cards, ls.deck_size))),
# ls.draw implies ls.discard (and converse true before the ls.extraround)
Implies(ls.draw_any[m], ls.discard_any[m]),
Implies(ls.discard_any[m], Or(ls.extraround[m], ls.draw_any[m])),
# ls.play requires ls.discard
Implies(ls.play[m], ls.discard_any[m]),
# definition of ls.play5
Iff(ls.play5[m], And(ls.play[m], Or(ls.discard[m][i] for i in range(ls.deck_size) if deck[i].rank == 5))),
# definition of ls.incr_clues
Iff(ls.incr_clues[m], And(ls.discard_any[m], Implies(ls.play[m], And(ls.play5[m], Not(ls.clues[m-1][8]))))),
# change of ls.clues
*[Iff(ls.clues[m][i], Or(ls.clues[m-1][i+1], And(ls.clues[m-1][i], Or(ls.discard_any[m], ls.dummyturn[m])), And(ls.clues[m-1][i-1], ls.incr_clues[m]))) for i in range(1, 9)],
## more than 8 clues not allowed, ls.discarding produces a strike
# Note that this means that we will never strike while not at 8 clues.
# It's easy to see that if there is any solution to the instance, then there is also one where we only strike at 8 clues
# (or not at all) -> Just strike later if neccessary
# So, we decrease the solution space with this formulation, but do not change whether it's empty or not
Iff(ls.strike[m], And(ls.discard_any[m], Not(ls.play[m]), ls.clues[m-1][8])),
# change of strikes
*[Iff(ls.strikes[m][i], Or(ls.strikes[m-1][i], And(ls.strikes[m-1][i-1], ls.strike[m]))) for i in range(1, ls.num_strikes + 1)],
# less than 0 clues not allowed
Implies(Not(ls.discard_any[m]), Or(ls.clues[m-1][1], ls.dummyturn[m])),
# we can only draw card i if the last ls.drawn card was i-1
*[Implies(ls.draw[m][i], Or(And(ls.draw[m0][i-1], *[Not(ls.draw_any[m1]) for m1 in range(m0+1, m)]) for m0 in range(max(-1, m-9), m))) for i in range(ls.distributed_cards, ls.deck_size)],
# we can only draw at most one card (NOTE: redundant, FIXME: avoid quadratic formula)
AtMostOne(ls.draw[m][i] for i in range(ls.distributed_cards, ls.deck_size)),
# we can only discard a card if we drew it earlier...
*[Implies(ls.discard[m][i], Or(ls.draw[m0][i] for m0 in range(m-ls.num_players, -1, -ls.num_players))) for i in range(ls.distributed_cards, ls.deck_size)],
# ...or if it was part of the initial hand
*[Not(ls.discard[m][i]) for i in range(0, ls.distributed_cards) if i // ls.hand_size != m % ls.num_players],
# we can only discard a card if we did not discard it yet
*[Implies(ls.discard[m][i], And(Not(ls.discard[m0][i]) for m0 in range(m-ls.num_players, -1, -ls.num_players))) for i in range(ls.deck_size)],
# we can only discard at most one card (FIXME: avoid quadratic formula)
AtMostOne(ls.discard[m][i] for i in range(ls.deck_size)),
# we can only play a card if it matches the progress
*[Implies(
And(ls.discard[m][i], ls.play[m]),
And(
Not(ls.progress[m-1][deck[i].suitIndex, deck[i].rank]),
ls.progress[m-1][deck[i].suitIndex, deck[i].rank-1 ]
)
)
for i in range(ls.deck_size)
],
# change of progress
*[
Iff(
ls.progress[m][s, r],
Or(
ls.progress[m-1][s, r],
And(ls.play[m], Or(ls.discard[m][i]
for i in range(0, ls.deck_size)
if deck[i] == DeckCard(s, r) ))
)
)
for s in range(0, ls.num_suits)
for r in range(1, 6)
],
# extra round bool
Iff(ls.extraround[m], Or(ls.extraround[m-1], ls.draw[m-1][ls.deck_size-1])),
# dummy turn bool
*[Iff(ls.dummyturn[m], Or(ls.dummyturn[m-1], ls.draw[m-1 - ls.num_players][ls.deck_size-1])) for i in range(0,1) if m >= ls.num_players]
)
win = And(
# maximum progress at each color
*[ls.progress[ls.max_moves-1][s, 5] for s in range(0, ls.num_suits)],
# played every color/value combination (NOTE: redundant, but makes solving faster)
*[
Or(
And(ls.discard[m][i], ls.play[m])
for m in range(ls.max_moves)
for i in range(ls.deck_size)
if deck[i] == DeckCard(s, r)
)
for s in range(0, ls.num_suits)
for r in range(1, 6)
]
)
constraints = And(*[valid_move(m) for m in range(ls.max_moves)], win)
2023-03-02 11:03:33 +01:00
# print('Solving instance with {} variables, {} nodes'.format(len(get_atoms(constraints)), get_formula_size(constraints)))
model = get_model(constraints)
if model:
# print_model(model, deck)
solution = toJSON(model, deck, ls)
2023-02-28 22:33:06 +01:00
return True, solution
else:
print('unsatisfiable')
2023-02-28 22:33:06 +01:00
return False, None
#conj = list(conjunctive_partition(constraints))
#print('statements: {}'.format(len(conj)))
#ucore = get_unsat_core(conj)
#print('unsat core size: {}'.format(len(ucore)))
#for f in ucore:
# print(f.serialize())
2023-03-02 11:03:33 +01:00
def print_model(model, deck, num_players):
draw = globals()['draw'][num_players]
for m in range(max_moves[num_players]):
2023-02-28 14:19:01 +01:00
print('=== move {} ==='.format(m))
print('clues: ' + ''.join(str(i) for i in range(1, 9) if model.get_py_value(clues[m][i])))
2023-02-28 15:41:23 +01:00
print('strikes: ' + ''.join(str(i) for i in range(1, NUM_STRIKES) if model.get_py_value(strikes[m][i])))
2023-02-28 14:19:01 +01:00
print('draw: ' + ', '.join('{} [{}{}]'.format(i, deck[i][0], deck[i][1]) for i in range(20, 50) if model.get_py_value(draw[m][i])))
print('discard: ' + ', '.join('{} [{}{}]'.format(i, deck[i][0], deck[i][1]) for i in range(50) if model.get_py_value(discard[m][i])))
for c in COLORS:
2023-02-28 14:19:01 +01:00
print('progress {}: '.format(c) + ''.join(str(k) for k in range(1, 6) if model.get_py_value(progress[m][c, k])))
2023-02-28 19:09:13 +01:00
flags = ['discard_any', 'draw_any', 'play', 'play5', 'incr_clues', 'strike', 'extraround', 'dummyturn']
2023-02-28 14:19:01 +01:00
print(', '.join(f for f in flags if model.get_py_value(globals()[f][m])))
def toJSON(model, deck: List[DeckCard], ls: Literals) -> dict:
players = ["Alice", "Bob", "Cathy", "Donald", "Emily"][:ls.num_players]
# we keep track of the hands to output some dummy clues
hands = [deck[ls.hand_size * p : ls.hand_size *(p+1)] for p in range(0, ls.num_players)]
actions = []
for m in range(ls.max_moves):
if model.get_py_value(ls.dummyturn[m]):
break
if model.get_py_value(ls.discard_any[m]):
discarded = next(i for i in range(0,ls.deck_size) if model.get_py_value(ls.discard[m][i]))
icard = hands[m % ls.num_players].index(deck[discarded])
for i in range(icard, ls.hand_size - 1):
hands[m % ls.num_players][i] = hands[m % ls.num_players][i + 1]
if model.get_py_value(ls.draw_any[m]):
hands[m % ls.num_players][ls.hand_size - 1] = next(deck[i] for i in range(ls.distributed_cards, ls.deck_size) if model.get_py_value(ls.draw[m][i]))
if model.get_py_value(ls.play[m]) or model.get_py_value(ls.strike[m]):
actions.append( Action( ActionType.Play, target=discarded) )
else:
actions.append( Action( ActionType.Discard, target=discarded) )
else:
actions.append( Action(
ActionType.RankClue,
target=(m+1) % ls.num_players, # clue next player
value=hands[(m+1) % ls.num_players][0].rank # clue rank of rightmost card
)
)
actions.append( Action (ActionType.EndGame, target=0, value=1))
game = {
"deck": deck,
"players": players,
"actions": actions,
"first_player": 0,
"options": {
"variant": "No Variant",
}
}
return game
2023-02-28 14:19:01 +01:00
2023-03-02 11:03:33 +01:00
if __name__ == "__main__":
COLORS = 'rygbp'
deck_str = 'p5 p3 b4 r5 y4 y4 y5 r4 b2 y2 y3 g5 g2 g3 g4 p4 r3 b2 b3 b3 p4 b1 p2 b1 b1 p2 p1 p1 g1 r4 g1 r1 r3 r1 g1 r1 p1 b4 p3 g2 g3 g4 b5 y1 y1 y1 r2 r2 y2 y3'
deck = [DeckCard(COLORS.index(c[0]), int(c[1])) for c in deck_str.split(" ")]
print(deck)
solvable, sol = solve(deck, num_players=5)
2023-03-02 11:03:33 +01:00
if solvable:
print(sol)
print(link(sol))