if (not feasibleActions) or len(moveSeq)>wordLength:
break
print "Run: %d\t Iteration: %d\t Word: %d\t factor (%d|%d)" %(i, numIter, t, len(currFactors), len(allowedFactors))
while feasibleActions:
# If there are no feasible actions available
if not feasibleActions:
break
chosenAction = random.choice(list(feasibleActions))
tentativeMoveSeq = moveSeq+[chosenAction]
# Compute all k-subsequences
tempFactors = generateKSubsequences(tentativeMoveSeq, 2)
# Check if the factors generated are subset of allowed factors
if tempFactors.issubset(allowedFactors):
currState = getNextPosition(arenaDimensions, currState, chosenAction)
currFactors.update(tempFactors)
moveSeq = deepcopy(tentativeMoveSeq)
numIter += 1
numFactors.append(len(currFactors))
break
else:
feasibleActions.remove(chosenAction)
# Check if learning is complete or not
if currFactors == allowedFactors:
learningComplete = True
def computeSPkGrammar(self, kval, moveSeq):
'''Compute the SP-k grammar from the input move sequence'''
return generateKSubsequences(moveSeq, kval)
break
print "Run: %d\t Iteration: %d\t Word: %d\t factor (%d|%d)" % (
i, numIter, t, len(currFactors), len(allowedFactors))
while feasibleActions:
# If there are no feasible actions available
if not feasibleActions:
break
chosenAction = random.choice(list(feasibleActions))
tentativeMoveSeq = moveSeq + [chosenAction]
# Compute all k-subsequences
tempFactors = generateKSubsequences(tentativeMoveSeq, 2)
# Check if the factors generated are subset of allowed factors
if tempFactors.issubset(allowedFactors):
currState = getNextPosition(arenaDimensions, currState,
chosenAction)
currFactors.update(tempFactors)
moveSeq = deepcopy(tentativeMoveSeq)
numIter += 1
numFactors.append(len(currFactors))
break
else:
feasibleActions.remove(chosenAction)
# Check if learning is complete or not
if currFactors == allowedFactors: