return valueF
if __name__ == '__main__':
grid = negativeGrid()
valueF = {}
policy = {}
for s in grid.allStates():
valueF[s] = 0
# state -> action
policy = {
(2, 0): 'U',
(1, 0): 'U',
(0, 0): 'R',
(0, 1): 'R',
(0, 2): 'R',
(1, 2): 'R',
(2, 1): 'R',
(2, 2): 'R',
(2, 3): 'U',
}
statesRewardsList = []
printPolicy(policy, grid)
for n in range(2000):
valueF = playGame(grid, policy)
print("final values:")
printValues(valueF, grid)
printPolicy(policy, grid)
print("\n\n")
stateActionsReturnsList = []
printPolicy(policy, grid)
for n in range(2000):
#print(n)
seenStates = set()
stateActionsReturnsList = playGame(grid, policy)
for s, a, r in stateActionsReturnsList:
if (s, a) not in seenStates:
#first visit policy
seenStates.add((s, a))
if (s, a) in allReturns:
allReturns[(s, a)].append(r)
else:
allReturns[(s, a)] = r
valueF[s][a] = np.mean(allReturns[(s, a)])
#print (policy)
#print(valueF)
for s in states:
#print (valueF[s])
#print (getMaximumFromDict(valueF[s])[0])
policy[s] = getMaximumFromDict(valueF[s])[0]
V = {}
for s in policy.keys():
V[s] = getMaximumFromDict(valueF[s])[1]
print("final values:")
printValues(V, grid)
#print(valueF)
printPolicy(policy, grid)
print("\n\n")
