def get_p_k(k, ZTemp, T, lattice_size, Temp):
k -= 1 # converting to array indexes (was in range 1,..,lattice_size)
if k == lattice_size-1: # for p_last(y_last)
res_matrix = np.ndarray((2**lattice_size , 1), dtype=float)
# res_matrix = [0] * pow(2, lattice_size)
for y_last in range(pow(2, lattice_size)):
y_last_vector = y2row(y_last, width=lattice_size)
res_matrix[y_last] = (T[lattice_size-2][y_last] * G(y_last_vector, Temp)) / ZTemp
return res_matrix
elif k == 0: # p_(1|2)(y1|y2)
# res_matrix = [[0] * pow(2, lattice_size)] * pow(2, lattice_size)
res_matrix = np.ndarray((2**lattice_size, 2**lattice_size), dtype=float)
for y1 in range(pow(2, lattice_size)):
y1_vector = y2row(y1, width=lattice_size)
for y2 in range(pow(2, lattice_size)):
y2_vector = y2row(y2, width=lattice_size)
res_matrix[y1][y2] = (F(y1_vector, y2_vector, Temp) * G(y1_vector, Temp)) / T[0][y2]
return res_matrix
else:
# res_matrix = [[0] * pow(2, lattice_size)] * pow(2, lattice_size)
res_matrix = np.ndarray((2 ** lattice_size, 2 ** lattice_size), dtype=float)
for y_k in range(pow(2, lattice_size)):
y_k_vector = y2row(y_k, width=lattice_size)
for y_kplus1 in range(pow(2, lattice_size)):
y_kplus1_vector = y2row(y_kplus1, width=lattice_size)
res_matrix[y_k][y_kplus1] = (F(y_k_vector, y_kplus1_vector, Temp) * G(y_k_vector, Temp) * T[k-1][y_k]) / T[k][y_kplus1]
return res_matrix
def compute_efficient_3x3_Lattice_ZTemp(Temp):
y_s = [0, 1, 2, 3, 4, 5, 6, 7]
res = 0
for i in y_s:
for j in y_s:
for k in y_s:
y1 = y2row(i, width=3)
y2 = y2row(j, width=3)
y3 = y2row(k, width=3)
res += G(y1, Temp) * G(y2, Temp) * G(y3, Temp) * F(
y1, y2, Temp) * F(y2, y3, Temp)
return res
def calc_T(lattice_size, Temp, prev_T):
T_vec = [0]*pow(2, lattice_size)
res = 0
for y2 in range(pow(2, lattice_size)):
temp_sum = 0
for y1 in range(pow(2, lattice_size)):
y1_vector = y2row(y1, width=lattice_size)
y2_vector = y2row(y2, width=lattice_size)
temp_sum += G(y1_vector, Temp) * F(y1_vector, y2_vector, Temp)*prev_T[y1]
# res += temp_res
T_vec[y2] = temp_sum
return T_vec
def convert_y_to_image(y, lattice_size):
# image = [[]*lattice_size]*lattice_size
image = np.ndarray((lattice_size, lattice_size))
index = 0
for i in range(len(y)):
# row_values = [int(i) for i in list('{0:0b}'.format(row))]
row = y2row(y[i], width=lattice_size)
for j in range(len(row)):
if row[j] == 0:
row[j] = -1
# missing_zeros = lattice_size - len(row_values)
# padded_row_values = np.pad(row_values, (missing_zeros, 0), 'constant')
# image[index] = padded_row_values
# index += 1
image[i] = row
# np_image = np.array([arr for arr in image])
return image
def get_T_arrays(lattice_size, Temp):
print("Getting T:")
s = time.time()
T_arrays = [[]] * lattice_size
prev_T = [1] * pow(2, lattice_size) # T_0
for k in range(lattice_size-1):
curr_T = calc_T(lattice_size, Temp, prev_T)
T_arrays[k] = curr_T
# print('T' + str(k+1) + ': ' + str(curr_T))
prev_T = curr_T
# Printing the last T, the normalizing factor Z_temp
T_arrays[lattice_size - 1] = 0
for y_last in range(pow(2, lattice_size)):
y_last_vector = y2row(y_last, width=lattice_size)
T_arrays[lattice_size-1] += curr_T[y_last] * G(y_last_vector, Temp)
e = time.time()
print("Done (" + str(round(e - s, 2)) + " secs).")
return T_arrays