def main():
A = matrix.create_matrix(stdin.readline().split())
B = matrix.create_matrix(stdin.readline().split())
pi = matrix.create_matrix(stdin.readline().split())
emissions = matrix.create_vector(stdin.readline().split())
T = len(emissions) # nb of observations
N = len(A)
delta_matrix = [[0.0 for _ in range(N)] for _ in range(T)]
# remember that delta_id_matrix always have the index written as actual index - 2 (instead of -1)
# since t = 1 has no predecessor
delta_id_matrix = [[0 for _ in range(N)] for _ in range(T - 1)]
for i in range(N):
delta_matrix[0][i] = B[i][emissions[0]] * pi[0][i]
for t in range(1, T):
for i in range(N):
temp = []
for j in range(N):
temp.append(A[j][i] * delta_matrix[t - 1][j] *
B[i][emissions[t]])
delta_matrix[t][i] = max(temp)
delta_id_matrix[t - 1][i] = temp.index(delta_matrix[t][i])
seq = [delta_matrix[T - 1].index(max(delta_matrix[T - 1]))]
for t in reversed(range(T - 1)):
seq.append(delta_id_matrix[t][seq[-1]])
seq = reversed(seq)
result = ""
for state in seq:
result += str(state) + " "
print(result[:-1])
return result[:-1]
def main():
A = matrix.create_matrix(stdin.readline().split())
B = matrix.create_matrix(stdin.readline().split())
pi = matrix.create_matrix(stdin.readline().split())
v1 = matrix.matrix_vec_mul(pi, A)
v2 = matrix.matrix_vec_mul(v1, B)
n = len(v2[0])
result = "1 " + str(n)
for i in range(0, n):
result += " " + str(v2[0][i])
print(result)
return result
def linkingnumber(overstrandlist, signlist, colourlist, p):
dln_mat = []
for k in range((p - 1) // 2 + 1):
coeff_matrix = matrix.create_matrix(overstrandlist, colourlist, signlist, p, k)
coeff_dict = rrematrix_to_dict(coeff_matrix, p, k)
# print("k =", k)
# print("Coefficients:", coeff_dict)
# print()
dln = matrix_to_dln(colourlist, overstrandlist, signlist, coeff_dict, p, k)
dln_mat.append(dln)
return dln_mat
def main():
A = matrix.create_matrix(stdin.readline().split())
B = matrix.create_matrix(stdin.readline().split())
pi = matrix.create_matrix(stdin.readline().split())
emissions = matrix.create_vector(stdin.readline().split())
T = len(emissions) # nb of observations
N = len(A)
alpha_matrix = [[0.0 for _ in range(N)] for _ in range(T)]
for i in range(N):
alpha_matrix[0][i] = B[i][emissions[0]] * pi[0][i]
for t in range(1, T):
for i in range(N):
s = 0
for j in range(N):
s += A[j][i] * alpha_matrix[t - 1][j]
alpha_matrix[t][i] = B[i][emissions[t]] * s
result = 0
for i in range(N):
result += alpha_matrix[T - 1][i]
print(result)
return result
def gauss_to_dln(gauss, p):
overstrand = gauss_to_overstrand.create_overstrand_list(gauss)
signlist = gauss_to_signlist.SignList(gauss)
colourlists = rrematrix_to_colourlist.overstrand_to_colourlist(overstrand, p)
# print(signlist, overstrand, colourlists)
complete_dln = []
for i in range(len(colourlists)):
colourlist = colourlists[i]
if len(colourlist) % 2 == 1:
colourlist.append(colourlist[0])
if i == 0:
signlist.append(1)
overstrand.append(len(overstrand))
# print(signlist, overstrandlist, colourlist)
dln_mat = []
for k in range(((p - 1) // 2) + 1):
# print(k)
mat = matrix.create_matrix(overstrand, colourlist, signlist, p, k)
coeffs = rrematrix_to_dict(mat, p, k)
# print("Coefficients: k = ", k)
# print(coeffs)
# print()
dlns = matrix_to_dln(colourlist, overstrand, signlist, coeffs, p, k)
dln_mat.append(dlns)
complete_dln.append(dln_mat)
# print(complete_dln)
error_list = errors(complete_dln, p)
if error_list == []:
return complete_dln
else:
return ["Error:", error_list]
import matrix a = matrix.rand_matrix(4, 4) b = matrix.rand_matrix(4, 4) c = matrix.create_matrix(4, 4) matrix.mult(c, a, b) print c a = matrix.rand_matrix(4, 2) b = matrix.rand_matrix(2, 4) c = matrix.create_matrix(4, 4) c = matrix.mult(c, a, b) print c r = matrix.create_matrix(2, 2) print matrix.mult(r, [[1, 0], [0, 1]], [[1, 0], [0, 1]]) print matrix.mult(r, [[2, 0], [0, 2]], [[0, 2], [2, 0]])
import sys
import matrix
if __name__ == "__main__":
#exec(open("dataPreProcessing.py").read())
kmin = int(sys.argv[1])
kmax = int(sys.argv[2])
for k in range(kmin, kmax + 1):
print("Run for k = " + str(k))
matrix.create_matrix(k)
from gauss import gauss, residuo
from gauss_seidel import gauss_seidel, converges
from matrix import create_matrix, n
import numpy as np
A, b = create_matrix()
g_res = gauss(A, b)
g_mres = max(abs(residuo(A, b, g_res)))
print('A1:')
print('Primeira:', g_res[0])
print('Última:', g_res[-1])
print('Resíduo máximo:', g_mres)
print('\nA2:')
print('Operações em ponto flutuante:', (4 * n**3 + 9 * n**2 - n - 6) // 6)
criteria = 1e-4
print('\nB1:')
if converges(A):
print('O sistema possui diagonal dominante e sempre irá convergir. '
'Podemos acelerar o cálculo usando um fator de sobre-relaxação.')
else:
print('O sistema não possui diagonal dominante e '
'talvez não venha a convergir')
def matrix_to_dln(colourlist, overstrandlist, signlist, coeff_dict, p, k):
universes_list = i2l.universe_lists(colourlist, overstrandlist, p)
wheres = i2l.where_lists(colourlist, overstrandlist, p)
vert_order = i2l.vertical_order(colourlist, overstrandlist, p)
dlns = []
dln = 0
# print('K = ', k)
if coeff_dict == "X": # No DLN Exists
for i in range(((p - 1) // 2) + 1):
dlns.append("x")
else:
# Linking number of Index 1 knot with surface
zero_2chain = rrematrix_to_dict(matrix.create_matrix(overstrandlist, colourlist, signlist, p, 0), p, 0)
if zero_2chain == "X":
dln = "x"
else:
for i in range(len(colourlist)):
if colourlist[i] == colourlist[overstrandlist[i]]:
# print("homogeneous => no change. DLN = ", dln)
dln += 0
else:
over_index = vert_order[i][-1] # Need bottom overstrand
where_over = wheres[over_index - 1][overstrandlist[i]]
if where_over == colourlist[i]:
epsilon1 = 1
else:
epsilon1 = -1
if signlist[i] * epsilon1 == 1:
# Walk through R wall
dln += signlist[i] * coeff_dict[(over_index, overstrandlist[i])]
else:
# Walk through L wall
dln -= signlist[i] * coeff_dict[(over_index, overstrandlist[i])]
if k == over_index:
dln += signlist[i]
dlns.append(dln)
# Linking numbers of Index 2 knots with surface
for j in range(len(wheres)): # Go through each of the index 2 knots
where_list = wheres[j]
dln = 0
j_2chain = rrematrix_to_dict(matrix.create_matrix(overstrandlist, colourlist, signlist, p, j + 1), p, j + 1)
if j_2chain == "X":
dln = "x"
else:
for i in range(len(where_list) - 1): # Go through each crossing
where_under = where_list[i]
universes = universes_list[i]
# print("crossing:", i)
if colourlist[i] == colourlist[overstrandlist[i]]:
for x in range(len(universes)): # list of pairs of universes
if where_under in universes[x]:
over_index = vert_order[i][x]
else:
if where_under in universes[0] and where_under in universes[1]:
# Must be on the top level, so a(j) = 0
over_index = 0
else:
if where_under in universes[0]:
level = i2l.index(universes[0], where_under)
else:
level = i2l.index(universes[1], where_under)
over_index = vert_order[i][level - 1]
if over_index == 0:
where_over = colourlist[overstrandlist[i]]
# print("adding index 1 wall")
# print("coefficient:", coeff_dict[(0, overstrandlist[i])])
dln += signlist[i] * coeff_dict[(0, overstrandlist[i])]
# print()
# print(dln)
else:
where_over = wheres[over_index - 1][overstrandlist[i]]
# print(where_under, where_over)
reflections = [dln5.reflect(where_over, colourlist[i], p),
dln5.reflect(where_over, colourlist[overstrandlist[i]], p)]
if where_under in reflections:
epsilon1 = 1
else:
epsilon1 = -1
# print("adding wall", (over_index, overstrandlist[i]))
# if epsilon1 * signlist[i] == -1:
# print("right")
# else:
# print("left")
# print("coefficient:", -epsilon1 * signlist[i] * coeff_dict[(over_index, overstrandlist[i])])
dln -= epsilon1 * coeff_dict[(over_index, overstrandlist[i])]
if k == over_index:
# print("since over_index = k, add", (signlist[i] + epsilon1) // 2)
dln += (signlist[i] + epsilon1) // 2
# print()
# print("Final DLN:", dln)
# print()
dlns.append(dln)
return dlns
# Print the time.txt in sort order
results = []
# This opens a handle to your file, in 'r' read mode
with open('time.txt') as fp:
# Read in all the lines of your file into a list of lines
for line in fp:
results.append(line)
float_results = [float(x) for x in results]
data = float_results[1::2]
y = np.array(float_results[0::2])
y = y.astype(int)
dictionary = {}
dictionary = dict(zip(y, data))
sort_dictionary = dict(sorted(dictionary.items(), key=operator.itemgetter(0)))
data_x = np.cumsum(np.array(list(sort_dictionary.values())))
# Plot time of k files
plt.plot(data_x, sort_dictionary.keys(), color="blue")
plt.ylabel('number of k files')
plt.xlabel('time(sec)')
plt.title('Time plot')
plt.savefig('time.png')
plt.show()
matrix.create_matrix(sys.argv[2], best_k)
def main():
A = matrix.create_matrix(stdin.readline().split())
B = matrix.create_matrix(stdin.readline().split())
pi = matrix.create_matrix(stdin.readline().split())
emissions = matrix.create_vector(stdin.readline().split())
T = len(emissions)
N = len(A)
K = len(B[0])
alpha = [[0.0 for _ in range(N)] for _ in range(T)]
beta = [[0.0 for _ in range(N)] for _ in range(T)]
di_gamma = [[[0.0 for _ in range(N)] for _ in range(N)]
for _ in range(T - 1)]
gamma = [[0.0 for _ in range(N)] for _ in range(T)]
scaling_factors = [0.0 for _ in range(T)]
max_its = 30
its = 0
old_log_prob = -float("inf")
range_n = range(N)
while its < max_its: # its < max_its:
scaling_factors[0] = 0
for i in range_n:
temp = B[i][emissions[0]] * pi[0][i]
alpha[0][i] = temp
scaling_factors[0] += temp
scaling_factors[0] = 1 / scaling_factors[0]
alpha[0] = [x * scaling_factors[0] for x in alpha[0]]
log_prob = log(scaling_factors[0])
for t in range(1, T):
scaling_factors[t] = 0
for i in range_n:
s = 0
for j in range_n:
s += A[j][i] * alpha[t - 1][j]
alpha[t][i] = B[i][emissions[t]] * s
scaling_factors[t] += alpha[t][i]
scaling_factors[t] = 1 / scaling_factors[t]
alpha[t] = [x * scaling_factors[t] for x in alpha[t]]
log_prob += log(scaling_factors[t])
log_prob = -log_prob
if log_prob > old_log_prob:
old_log_prob = log_prob
# its += its
else:
break
beta[T - 1] = [scaling_factors[T - 1]
for _ in range_n] # 1 scaled by cT-1
for t in reversed(range(T - 1)):
for i in range(N):
s = 0
for j in range(N):
s += beta[t + 1][j] * B[j][emissions[t + 1]] * A[i][j]
beta[t][i] = s * scaling_factors[t]
for t in range(T - 1):
for i in range_n:
di_gamma_sum = 0
for j in range_n:
temp = alpha[t][i] * A[i][j] * B[j][emissions[
t + 1]] * beta[t + 1][j]
di_gamma[t][i][j] = temp
di_gamma_sum += temp
gamma[t][i] = di_gamma_sum
gamma[T - 1] = alpha[T - 1]
for i in range_n:
pi[0][i] = gamma[0][i]
gamma_sum = 0
for t in range(T - 1):
gamma_sum += gamma[t][i]
for k in range(K):
ind_gamma_sum = 0
for t in range(T):
if emissions[t] == k:
ind_gamma_sum += gamma[t][i]
B[i][k] = ind_gamma_sum / gamma_sum
for j in range_n:
di_gamma_sum = 0
for t in range(T - 1):
di_gamma_sum += di_gamma[t][i][j]
A[i][j] = di_gamma_sum / gamma_sum
its += 1
result = str(N) + " " + str(N)
for row in A:
for val in row:
result += " " + str(round(val, 6))
result += "\n" + str(N) + " " + str(K)
for row in B:
for val in row:
result += " " + str(round(val, 6))
print(result)
return result
from graph import Graph
from binary_heap import BinaryHeap
from binomial_heap import BinomialHeap
from dijkstra import dijkstra #(List[int],float) or Map
from prima import prima #(List[Tuple],float)
from alfa_star import a_star #(List[int],float)
from bellman_ford import bellman_ford #(List[int],float) or Map
from floyd_warshall import floyd_warshall #(List[int],float) or Map or
#(List[float],List[int])
from kruskal import kruskal #(List[Tuple],float)
# bubble: 1.23
G = Graph(*create_matrix(25,25,20,dst=7))
G.draw()
H = Graph(G.H())
q=H.get_queue()
lst=[]
while q:
lst.append(q.get())
def test(constructor, lst):
n=len(lst)
q=constructor()
b=constructor()
# =============================================================================
# try:
# asd = q._lst
