def transModel(s):
if s == 'a':
return dist.deltaDist('a')
elif s == 'z':
return dist.deltaDist('z')
else:
return dist.DDist({s:0.5, nextChar(s):0.25, prevChar(s):0.25})
def transModel(s):
distance = FORWARD_VELOCITY * 0.1
result1 = (float(s) / numStates) * (xMax - xMin) + distance
result2 = (float(s+1) / numStates) * (xMax - xMin) + distance
state1 = clip(int(float(result1) * numStates / (xMax - xMin)), 0, numStates-1)
state2 = clip(int(float(result2) * numStates / (xMax - xMin)), 0, numStates-1)
prob = float(distance) / (xMax - xMin) * numStates
return dist.mixture(dist.deltaDist(state1), dist.deltaDist(state2), 1.0 - (prob % 1.0))
def transModel(s):
distance = FORWARD_VELOCITY * 0.1
result1 = (float(s) / numStates) * (xMax - xMin) + distance
result2 = (float(s + 1) / numStates) * (xMax - xMin) + distance
state1 = clip(int(float(result1) * numStates / (xMax - xMin)), 0,
numStates - 1)
state2 = clip(int(float(result2) * numStates / (xMax - xMin)), 0,
numStates - 1)
prob = float(distance) / (xMax - xMin) * numStates
return dist.mixture(dist.deltaDist(state1), dist.deltaDist(state2),
1.0 - (prob % 1.0))
def obsModel(s):
if numObservations < 10:
return dist.mixture(dist.deltaDist(ideal[s]),
dist.uniformDist(range(numObservations)),
float(numObservations) / 15.)
return dist.mixture(
dist.triangleDist(ideal[s],
int(numObservations / 6.0) + 2),
dist.uniformDist(range(numObservations)), 0.8)
def teleportModel2(state):
current_d = dist.deltaDist(state)
random_d = dist.uniformDist(range(numStates))
return dist.mixture(current_d, random_d, 0.7)
def obsModelA(state):
zero_dist = dist.uniformDist([0])
return dist.mixture(dist.deltaDist(ideal[state]), zero_dist, 0.7)
def obsModelD(state):
ideal_d = dist.deltaDist(ideal[state])
other_d = dist.deltaDist(9 - ideal[state])
return dist.mixture(ideal_d, other_d, 0.5)
def teleportModel(state):
right_d = dist.deltaDist(min(state+1, numStates-1))
random_d = dist.uniformDist(range(numStates))
return dist.mixture(right_d, random_d, 0.7)
def perfectObsModel(state):
return dist.deltaDist(ideal[state])
def obsModelB(state):
random_dist = dist.squareDist(0, 10)
return dist.mixture(dist.deltaDist(ideal[state]), random_dist, 0.7)
def resetModel(state):
nominal = moveRightModel(state)
return dist.mixture(dist.deltaDist(0), nominal, 0.3)
def resetModel2(state):
nominal = dist.deltaDist(state)
return dist.mixture(dist.deltaDist(0), nominal, 0.3)
def perfectObsModel(state):
return dist.deltaDist(ideal[state])
def obsModel(s):
if numObservations < 10:
return dist.mixture(dist.deltaDist(ideal[s]), dist.uniformDist(range(numObservations)), float(numObservations) / 15.)
return dist.mixture(dist.triangleDist(ideal[s], int(numObservations / 6.0) + 2), dist.uniformDist(range(numObservations)), 0.8)
def naiveTransModel(s):
return dist.deltaDist(clip(s+1, 0, numStates-1))
def naiveObsModel(s):
return dist.deltaDist(ideal[s])
return dist.uniformDist(range(numObservations))
def obsModel(s):
pass # your code here
def naiveTransModel(s):
return dist.deltaDist(clip(s+1, 0, numStates-1))
transModel = models.makeTransModel(FORWARD_VELOCITY, xMax, xMin, numStates)
def confidentLocation(belief):
return (-1,False) #your code here
uniformInitDist = dist.squareDist(0, numStates)
spikedInitDist = dist.deltaDist(24)
INIT_DIST_TO_USE = uniformInitDist
OBS_MODEL_TO_USE = naiveObsModel
TRANS_MODEL_TO_USE = naiveTransModel
######################################################################
###
### Brain Methods -- do not change the following code
###
######################################################################
# Robot's Ideal Readings
ideal = idealReadings.computeIdealReadings(WORLD_FILE, xMin, xMax, robotY, numStates, numObservations)
def getParkingSpot(ideal):
def resetModel2(state):
nominal = dist.deltaDist(state)
return dist.mixture(dist.deltaDist(0), nominal, 0.3)
def teleportModel2(state):
nominal = dist.deltaDist(state)
return dist.mixture(dist.uniformDist(range(numStates)), nominal, 0.3)
def resetModel(state):
nominal = moveRightModel(state)
return dist.mixture(dist.deltaDist(0), nominal, 0.3)
def obsModelB(state):
random_dist = dist.squareDist(0, 10)
return dist.mixture(dist.deltaDist(ideal[state]), random_dist, 0.7)
def teleportModel2(state):
nominal = dist.deltaDist(state)
return dist.mixture(dist.uniformDist(range(numStates)), nominal,0.3)
def obsModelC(state):
return dist.deltaDist(9 - ideal[state])
def transModel(s):
r = (FORWARD_VELOCITY / 10.0) * numStates / (xMax - xMin)
return dist.mixture(dist.deltaDist(clip(s + int(r), 0, numStates - 1)),
dist.deltaDist(clip(s + int(r) + 1, 0, numStates - 1)),
int(r) + 1 - r)
def obsModelD(state):
ideal_d = dist.deltaDist(ideal[state])
other_d = dist.deltaDist(9 - ideal[state])
return dist.mixture(ideal_d, other_d, 0.5)
def obsModelA(state):
zero_dist = dist.uniformDist([0])
return dist.mixture(dist.deltaDist(ideal[state]), zero_dist, 0.7)
def moveRightModel(state):
return dist.deltaDist(min(state + 1, numStates - 1))
def obsModelC(state):
return dist.deltaDist(9 - ideal[state])
def teleportModel(state):
right_d = dist.deltaDist(min(state + 1, numStates - 1))
random_d = dist.uniformDist(range(numStates))
return dist.mixture(right_d, random_d, 0.7)
def moveRightModel(state):
return dist.deltaDist(min(state+1, numStates-1))
def resetModel(state):
right_d = dist.deltaDist(min(state + 1, numStates - 1))
zero_d = dist.deltaDist(0)
return dist.mixture(right_d, zero_d, 0.7)
def resetModel(state):
right_d = dist.deltaDist(min(state+1, numStates-1))
zero_d = dist.deltaDist(0)
return dist.mixture(right_d, zero_d, 0.7)
def teleportModel2(state):
current_d = dist.deltaDist(state)
random_d = dist.uniformDist(range(numStates))
return dist.mixture(current_d, random_d, 0.7)
def resetModel2(state):
current_d = dist.deltaDist(state)
zero_d = dist.deltaDist(0)
return dist.mixture(current_d, zero_d, 0.7)
def resetModel2(state):
current_d = dist.deltaDist(state)
zero_d = dist.deltaDist(0)
return dist.mixture(current_d, zero_d, 0.7)
def obsModelA(state):
return dist.mixture(dist.deltaDist(ideal[state]), dist.deltaDist(0), 0.7)
def obsModelB(state):
return dist.mixture(dist.deltaDist(ideal[state]),dist.uniformDist(range(10)),0.7)
def obsModelA(state):
return dist.mixture(dist.deltaDist(ideal[state]),dist.deltaDist(0),0.7)
def obsModelB(state):
return dist.mixture(dist.deltaDist(ideal[state]),
dist.uniformDist(range(10)), 0.7)
