def update(self, solved_prob, rewards):
for i, reward in enumerate(rewards):
if reward >= 0:
reward = reward / self.p[i]
self.w[i] = self.w[i] * np.exp(self.gamma * reward / len(SOLVERS))
#TODO: Implement pruning
if is_solved(solved_prob.result):
self.solved.append(solved_prob)
def update(self, solved_prob, rewards):
for i, r in enumerate(rewards):
if r > 0:
self.dist.update(i, 1)
elif r == 0:
self.dist.update(i, 0)
else:
pass
if is_solved(solved_prob.result):
self.solved.append(solved_prob)
def update(self, solved_prob, rewards):
X = np.array(solved_prob.datapoint).reshape(1, -1)
y = np.array([list(SOLVERS.keys()).index(solved_prob.solve_method)])
if self.fitted:
self.clf.partial_fit(X, y)
else:
self.clf.partial_fit(X, y, classes=np.unique(list(range(len(SOLVERS)))))
self.fitted = True
#TODO: Implement pruning
if is_solved(solved_prob.result):
self.solved.append(solved_prob)
def update(self, solved_prob, rewards):
for i, r in enumerate(rewards):
if r > 0:
self.totals[i] += r
self.counts[i] += 1
self.values[i] = self.totals[i] / self.counts[i]
elif r == 0:
self.counts[i] += 1
self.values[i] = self.totals[i] / self.counts[i]
else:
pass
if is_solved(solved_prob.result):
self.solved.append(solved_prob)
def update(self, solved_prob, rewards):
point = solved_prob.datapoint.reshape((len(solved_prob.datapoint), 1))
for i, r in enumerate(rewards):
if r >= 0:
self.A_0 += self.Cs[i].T @ np.linalg.inv(self.As[i]) @ self.Cs[i]
self.B_0 += self.Cs[i].T @ np.linalg.inv(self.As[i]) @ self.Bs[i]
self.As[i] = self.As[i] + point @ point.T
self.Bs[i] = self.Bs[i] + r * point
self.Cs[i] = self.Cs[i] + point @ point.T
self.A_0 += point @ point.T - self.Cs[i].T @ np.linalg.inv(self.As[i]) @ self.Cs[i]
self.B_0 += r * point - self.Cs[i].T @ np.linalg.inv(self.As[i]) @ self.Bs[i]
#TODO: Implement pruning
if is_solved(solved_prob.result):
self.solved.append(solved_prob)
def update(self, solved_prob, rewards):
#TODO: Implement pruning
if is_solved(solved_prob.result):
self.solved.append(solved_prob)