2º Trabalho avaliativo Algebra Linear Computacional

Eduardo de Castro Luiz - 20190066565

Solução com o decomposição em Valores Singulares (Pseudo-Inversa)

b) Faça um gráfico de uma função que retorna o número de condicionamento e o posto da matriz U, sendo o eixo-x a variação de r_ij e o eixo-y o valor
do número de condicionamento da matriz U. Escolha valores para K no intervalo (1, 200]
In [1]: import numpy as np

def distance(p1, p2):

return np.linalg.norm(p1 - p2)

def matrix_U(p1, p2):

r = distance(p1, p2)
if r == 0:
return 0
return 1 / (2 * np.pi) * np.log(1 / r)

def matrix_rank(matrix):
return np.linalg.matrix_rank(matrix)

def condition_number(matrix):
return np.linalg.cond(matrix)

import matplotlib.pyplot as plt

# Valores de k fornecidos
k_values = [1.05, 1.1, 1.2, 1.3, 1.5, 1.8, 2.0, 2.3, 2.6, 3, 3.5, 4, 5, 6.5, 8, 10, 20, 50, 100, 200]

# Matriz retangular U, posto da matriz e número de condicionamento para cada valor de k

matrix_rank_values = []
num_conditioning = []

# Conjunto de pontos P1
t1_values = np.linspace(0, 2 * np.pi, 60)
P1 = np.array([(np.cos(t), np.sin(t)) for t in t1_values])

# Conjunto de pontos P2
t2_values = np.linspace(0, 2 * np.pi, 30)

for k in k_values:
U_matrix = np.zeros((60, 30))

for i in range(60):
for j in range(30):
P2_point = k * np.array([np.cos(t2_values[j]), np.sin(t2_values[j])])
U_matrix[i, j] = matrix_U(P1[i], P2_point)

rank = matrix_rank(U_matrix)
matrix_rank_values.append(rank)

num_cond = condition_number(U_matrix)
num_conditioning.append(num_cond)

# Plotagem do gráfico
plt.figure(figsize=(10, 6))
plt.plot(k_values, matrix_rank_values, marker='o', label='Posto da Matriz U')
plt.plot(k_values, num_conditioning, marker='o', label='Número de Condicionamento')
plt.xlabel('Valores de k')
plt.ylabel('Posto / Número de Condicionamento')
plt.title('Posto e Número de Condicionamento da Matriz U para diferentes valores de k')
plt.legend()
plt.grid(True)
plt.show()
c) Monte um vetor com 60 posições da forma: b = b_i = sen(3t_1)
In [2]: import numpy as np

# Sequência de ângulos t1
t1_values = np.linspace(0, 2 * np.pi, 60)

# Vetor b
b = np.sin(3 * t1_values)

print(b)
[ 0.00000000e+00 3.14076712e-01 5.96367359e-01 8.18302776e-01
9.57422038e-01 9.99645611e-01 9.40700267e-01 7.86551556e-01
5.52800065e-01 2.63102564e-01 -5.32221748e-02 -3.64160575e-01
-6.38244184e-01 -8.47734428e-01 -9.71429893e-01 -9.96812007e-01
-9.21311978e-01 -7.52570770e-01 -5.07665800e-01 -2.11382624e-01
1.06293486e-01 4.13212186e-01 6.78311836e-01 8.74763085e-01
9.82684125e-01 9.91152831e-01 8.99312130e-01 7.16456740e-01
4.61092501e-01 1.59063496e-01 -1.59063496e-01 -4.61092501e-01
-7.16456740e-01 -8.99312130e-01 -9.91152831e-01 -9.82684125e-01
-8.74763085e-01 -6.78311836e-01 -4.13212186e-01 -1.06293486e-01
2.11382624e-01 5.07665800e-01 7.52570770e-01 9.21311978e-01
9.96812007e-01 9.71429893e-01 8.47734428e-01 6.38244184e-01
3.64160575e-01 5.32221748e-02 -2.63102564e-01 -5.52800065e-01
-7.86551556e-01 -9.40700267e-01 -9.99645611e-01 -9.57422038e-01
-8.18302776e-01 -5.96367359e-01 -3.14076712e-01 -7.34788079e-16]
d) Resolva o sistema linear, 60 x 30. Uv = b com a decomposição em valores singulares
In [3]: import numpy as np

def distance(p1, p2):

return np.linalg.norm(p1 - p2)

def matrix_U(p1, p2):

r = distance(p1, p2)
if r == 0:
return 0
return 1 / (2 * np.pi) * np.log(1 / r)

def solve_system(U_matrix, b_vector):

U_pseudo_inv = np.linalg.pinv(U_matrix)
v = U_pseudo_inv @ b_vector
return v

def phi_exata(r, theta):

return r**3 * np.sin(3 * theta)

# Conjunto de pontos P1
t1_values = np.linspace(0, 2 * np.pi, 60)
P1 = np.array([(np.cos(t), np.sin(t)) for t in t1_values])
 # Valores de theta_i para cada ponto P1
theta_values = np.arctan2(P1[:, 1], P1[:, 0])

# Vetor b
b = np.sin(3 * t1_values)

# Valores de k para análise de sensibilidade

k_values = [1.05, 1.1, 1.2, 1.3, 1.5, 1.8, 2.0, 2.3, 2.6, 3, 3.5, 4, 5, 6.5, 8, 10, 20, 50, 100, 200]

# Armazenar as soluções para diferentes valores de k

solutions = {}

for k in k_values:
# Conjunto de pontos P2 para o valor atual de k
t2_values = np.linspace(0, 2 * np.pi, 30)
P2 = np.array([k * np.array([np.cos(t), np.sin(t)]) for t in t2_values])

# Criar a matriz U para o valor atual de k

U_matrix = np.zeros((60, 30))

for i in range(60):
for j in range(30):
U_matrix[i, j] = matrix_U(P1[i], P2[j])

# Resolver o sistema linear para obter o vetor v

v = solve_system(U_matrix, b)

# Armazenar a solução para o valor atual de k

solutions[k] = v

# Imprimir os resultados para cada valor de k

for k, v in solutions.items():
print(f"Resultados para k = {k}:")
print(f"Vetor v: {v}")
print()
Resultados para k = 1.05:
Vetor v: [-3.33066907e-16 8.98573917e-01 1.43320199e+00 1.38634794e+00
7.75131838e-01 -1.54235490e-01 -1.02465021e+00 -1.48052927e+00
-1.33322781e+00 -6.39795499e-01 3.18355082e-01 1.14946034e+00
1.51202263e+00 1.25587518e+00 4.84934516e-01 -4.84934516e-01
-1.25587518e+00 -1.51202263e+00 -1.14946034e+00 -3.18355082e-01
6.39795499e-01 1.33322781e+00 1.48052927e+00 1.02465021e+00
1.54235490e-01 -7.75131838e-01 -1.38634794e+00 -1.43320199e+00
-8.98573917e-01 -1.33226763e-15]

Resultados para k = 1.1:

Vetor v: [-1.33226763e-15 1.04585912e+00 1.66545804e+00 1.60617856e+00
8.92011582e-01 -1.86102194e-01 -1.18870601e+00 -1.70688895e+00
-1.52906181e+00 -7.27423162e-01 3.71269755e-01 1.31885389e+00
1.72860591e+00 1.43315636e+00 5.52929688e-01 -5.52929688e-01
-1.43315636e+00 -1.72860591e+00 -1.31885389e+00 -3.71269755e-01
7.27423162e-01 1.52906181e+00 1.70688895e+00 1.18870601e+00
1.86102194e-01 -8.92011582e-01 -1.60617856e+00 -1.66545804e+00
-1.04585912e+00 -1.99840144e-15]

Resultados para k = 1.2:

Vetor v: [-9.76996262e-15 1.35941172e+00 2.16444056e+00 2.08678570e+00
1.15811313e+00 -2.42856267e-01 -1.54479108e+00 -2.21674383e+00
-1.98467896e+00 -9.43230957e-01 4.82885605e-01 1.71207898e+00
2.24306266e+00 1.85929004e+00 7.17268050e-01 -7.17268050e-01
-1.85929004e+00 -2.24306266e+00 -1.71207898e+00 -4.82885605e-01
9.43230957e-01 1.98467896e+00 2.21674383e+00 1.54479108e+00
2.42856267e-01 -1.15811313e+00 -2.08678570e+00 -2.16444056e+00
-1.35941172e+00 -1.15463195e-14]

Resultados para k = 1.3:

Vetor v: [-1.77635684e-14 1.72839690e+00 2.75192967e+00 2.65318746e+00
1.47243867e+00 -3.08791075e-01 -1.96409165e+00 -2.81840853e+00
-2.52333934e+00 -1.19921730e+00 6.13962308e-01 2.17675966e+00
2.85184423e+00 2.36390707e+00 9.11935719e-01 -9.11935719e-01
-2.36390707e+00 -2.85184423e+00 -2.17675966e+00 -6.13962308e-01
1.19921730e+00 2.52333934e+00 2.81840853e+00 1.96409165e+00
3.08791075e-01 -1.47243867e+00 -2.65318746e+00 -2.75192967e+00
-1.72839690e+00 -2.48689958e-14]

Resultados para k = 1.5:

Vetor v: [-1.13686838e-13 2.65513932e+00 4.22747600e+00 4.07578934e+00
2.26193926e+00 -4.74361017e-01 -3.01721029e+00 -4.32959937e+00
-3.87631777e+00 -1.84222003e+00 9.43160586e-01 3.34390724e+00
4.38096214e+00 3.63139993e+00 1.40090246e+00 -1.40090246e+00
-3.63139993e+00 -4.38096214e+00 -3.34390724e+00 -9.43160586e-01
 1.84222003e+00 3.87631777e+00 4.32959937e+00 3.01721029e+00
4.74361017e-01 -2.26193926e+00 -4.07578934e+00 -4.22747600e+00
-2.65513932e+00 -8.52651283e-14]

Resultados para k = 1.8:

Vetor v: [-1.36424205e-12 4.58808074e+00 7.30507852e+00 7.04296397e+00
3.90863104e+00 -8.19695839e-01 -5.21373939e+00 -7.48154770e+00
-6.69827711e+00 -3.18335622e+00 1.62978149e+00 5.77827171e+00
7.57030258e+00 6.27505907e+00 2.42075945e+00 -2.42075945e+00
-6.27505907e+00 -7.57030258e+00 -5.77827171e+00 -1.62978149e+00
3.18335622e+00 6.69827711e+00 7.48154770e+00 5.21373939e+00
8.19695839e-01 -3.90863104e+00 -7.04296397e+00 -7.30507852e+00
-4.58808074e+00 -4.54747351e-13]

Resultados para k = 2.0:

Vetor v: [ 5.45696821e-12 6.29366356e+00 1.00206838e+01 9.66113028e+00
5.36163380e+00 -1.12441130e+00 -7.15190588e+00 -1.02627541e+01
-9.18830879e+00 -4.36674378e+00 2.23563991e+00 7.92629864e+00
1.03845029e+01 8.60776279e+00 3.32065768e+00 -3.32065768e+00
-8.60776279e+00 -1.03845029e+01 -7.92629864e+00 -2.23563991e+00
4.36674378e+00 9.18830879e+00 1.02627541e+01 7.15190588e+00
1.12441130e+00 -5.36163380e+00 -9.66113028e+00 -1.00206838e+01
-6.29366357e+00 4.54747351e-12]

Resultados para k = 2.3:

Vetor v: [-1.01863407e-10 9.57187557e+00 1.52402075e+01 1.46933715e+01
8.15437480e+00 -1.71008904e+00 -1.08771549e+01 -1.56083661e+01
-1.39742691e+01 -6.64127145e+00 3.40012885e+00 1.20549094e+01
1.57935308e+01 1.30913312e+01 5.05030525e+00 -5.05030525e+00
-1.30913312e+01 -1.57935308e+01 -1.20549094e+01 -3.40012885e+00
6.64127145e+00 1.39742691e+01 1.56083661e+01 1.08771549e+01
1.71008904e+00 -8.15437480e+00 -1.46933715e+01 -1.52402075e+01
-9.57187557e+00 -1.09139364e-10]

Resultados para k = 2.6:

Vetor v: [-6.69388101e-10 1.38271789e+01 2.20154424e+01 2.12255032e+01
1.17795095e+01 -2.47033163e+00 -1.57127372e+01 -2.25472707e+01
-2.01867144e+01 -9.59373608e+00 4.91170088e+00 1.74140781e+01
2.28147528e+01 1.89112548e+01 7.29548493e+00 -7.29548493e+00
-1.89112548e+01 -2.28147528e+01 -1.74140781e+01 -4.91170088e+00
9.59373608e+00 2.01867144e+01 2.25472707e+01 1.57127372e+01
2.47033163e+00 -1.17795095e+01 -2.12255032e+01 -2.20154424e+01
-1.38271789e+01 -6.40284270e-10]

Resultados para k = 3:
Vetor v: [ 8.38190317e-09 2.12411145e+01 3.38198080e+01 3.26063147e+01
1.80955141e+01 -3.79488815e+00 -2.41376823e+01 -3.46367949e+01
-3.10105422e+01 -1.47377603e+01 7.54528469e+00 2.67512579e+01
3.50476971e+01 2.90511994e+01 1.12072197e+01 -1.12072197e+01
-2.90511994e+01 -3.50476971e+01 -2.67512579e+01 -7.54528468e+00
1.47377602e+01 3.10105422e+01 3.46367949e+01 2.41376824e+01
3.79488813e+00 -1.80955141e+01 -3.26063147e+01 -3.38198080e+01
-2.12411145e+01 7.68341124e-09]

Resultados para k = 3.5:

Vetor v: [-8.56816769e-08 3.37301033e+01 5.37046023e+01 5.17776202e+01
2.87350060e+01 -6.02614170e+00 -3.83297457e+01 -5.50019474e+01
-4.92435926e+01 -2.34030173e+01 1.19816325e+01 4.24800070e+01
5.56544447e+01 4.61322290e+01 1.77966495e+01 -1.77966495e+01
-4.61322290e+01 -5.56544447e+01 -4.24800070e+01 -1.19816325e+01
2.34030173e+01 4.92435925e+01 5.50019475e+01 3.83297456e+01
6.02614188e+00 -2.87350062e+01 -5.17776200e+01 -5.37046026e+01
-3.37301030e+01 -8.94069672e-08]

Resultados para k = 4:
Vetor v: [-2.92062759e-06 5.03493142e+01 8.01654653e+01 7.72890473e+01
4.28930658e+01 -8.99528620e+00 -5.72152508e+01 -8.21020290e+01
-7.35064736e+01 -3.49339471e+01 1.78851161e+01 6.34103923e+01
8.30760196e+01 6.88621055e+01 2.65652585e+01 -2.65652589e+01
-6.88621044e+01 -8.30760215e+01 -6.34103896e+01 -1.78851196e+01
3.49339515e+01 7.35064683e+01 8.21020353e+01 5.72152435e+01
8.99529455e+00 -4.28930752e+01 -7.72890370e+01 -8.01654764e+01
-5.03493027e+01 -2.93552876e-06]

Resultados para k = 5:
Vetor v: [-1.37090683e-04 9.83387475e+01 1.56572957e+02 1.50955357e+02
8.37753637e+01 -1.75687935e+01 -1.11748635e+02 -1.60355449e+02
-1.43567387e+02 -6.82303278e+01 3.49318483e+01 1.23848421e+02
1.62257874e+02 1.34496255e+02 5.18853389e+01 -5.18853632e+01
-1.34496180e+02 -1.62257994e+02 -1.23848258e+02 -3.49320490e+01
6.82305642e+01 1.43567118e+02 1.60355755e+02 1.11748291e+02
1.75691793e+01 -8.37757950e+01 -1.50954882e+02 -1.56573471e+02
-9.83382092e+01 -1.37567520e-04]

Resultados para k = 6.5:

Vetor v: [-1.37329102e-03 2.16053085e+02 3.43987427e+02 3.31652847e+02
1.84049973e+02 -3.85935898e+01 -2.45517311e+02 -3.52294930e+02
-3.15423923e+02 -1.49895355e+02 7.67384567e+01 2.72101875e+02
3.56473709e+02 2.95494972e+02 1.13985550e+02 -1.13985821e+02
-2.95494217e+02 -3.56474922e+02 -2.72100304e+02 -7.67403450e+01
1.49897533e+02 3.15421551e+02 3.52297600e+02 2.45514198e+02
3.85971756e+01 -1.84054092e+02 -3.31648209e+02 -3.43992523e+02
-2.16047607e+02 -1.38092041e-03]

Resultados para k = 8:
Vetor v: [ 3.37524414e-01 4.02181641e+02 6.41852783e+02 6.17876221e+02
3.43486572e+02 -7.22136230e+01 -4.57554321e+02 -6.56910034e+02
-5.88019409e+02 -2.79452393e+02 1.43019165e+02 5.07380249e+02
6.64480286e+02 5.51054016e+02 2.12334595e+02 -2.12301147e+02
-5.51156189e+02 -6.64301575e+02 -5.07643921e+02 -1.42657471e+02
2.78981201e+02 5.88610962e+02 6.56191040e+02 4.58405762e+02
7.12321777e+01 -3.42384766e+02 -6.19083984e+02 -6.40561523e+02
-4.03526855e+02 3.43872070e-01]

Resultados para k = 10:

Vetor v: [ 1.33398438 784.46777344 1254.00830078 1207.18164062
669.67431641 -139.11791992 -896.1484375 -1280.20019531
-1151.38085938 -543.10693359 277.11230469 992.48779297
1297.18603516 1075.93481445 416.01293945 -416.79321289
-1073.67211914 -1300.70458984 -988.06542969 -281.99316406
547.95703125 1147.05517578 1283.53759766 894.18261719
139.45556641 -668.27587891 -1210.22949219 -1249.59472656
-789.78125 1.48095703]

Resultados para k = 20:

Vetor v: [-9.93188477e+00 6.30210596e+03 1.00296667e+04 9.64402417e+03
5.37115942e+03 -1.11801440e+03 -7.16834741e+03 -1.02507769e+04
-9.18544263e+03 -4.38183838e+03 2.24977393e+03 7.92544843e+03
1.03712686e+04 8.62360693e+03 3.31606238e+03 -3.33146582e+03
-8.59089697e+03 -1.03926056e+04 -7.93424072e+03 -2.21850415e+03
4.35556445e+03 9.18344873e+03 1.02795461e+04 7.13808350e+03
1.12276392e+03 -5.34577417e+03 -9.67695557e+03 -1.00196358e+04
-6.27784192e+03 -1.02520752e+01]

Resultados para k = 50:

Vetor v: [ 2.88325310e+00 9.83214384e+04 1.56567040e+05 1.50970982e+05
8.37834425e+04 -1.75837608e+04 -1.11758113e+05 -1.60341850e+05
-1.43556218e+05 -6.82427325e+04 3.49194359e+04 1.23859319e+05
1.62271411e+05 1.34486955e+05 5.18707970e+04 -5.18775464e+04
-1.34481082e+05 -1.62263949e+05 -1.23864189e+05 -3.49274449e+04
6.82465557e+04 1.43564582e+05 1.60339084e+05 1.11749559e+05
1.75854718e+04 -8.37748055e+04 -1.50971579e+05 -1.56575693e+05
-9.83228763e+04 2.88267326e+00]

Resultados para k = 100:

Vetor v: [-6.07085419e+00 7.86999410e+05 1.25275024e+06 1.20743947e+06
6.69931936e+05 -1.40496381e+05 -8.93686883e+05 -1.28273797e+06
-1.14878226e+06 -5.46079300e+05 2.79572057e+05 9.91085477e+05
1.29810495e+06 1.07569515e+06 4.14893568e+05 -4.14907731e+05
-1.07568869e+06 -1.29808742e+06 -9.91082193e+05 -2.79587835e+05
5.46067514e+05 1.14879156e+06 1.28275464e+06 8.93686797e+05
1.40479714e+05 -6.69940994e+05 -1.20742800e+06 -1.25273445e+06
-7.87000236e+05 -6.06922913e+00]

Resultados para k = 200:

Vetor v: [-2.26890671e+02 6.29424819e+06 1.00215061e+07 9.66134900e+06
5.36103302e+06 -1.12519013e+06 -7.15203822e+06 -1.02621033e+07
-9.18756779e+06 -4.36669923e+06 2.23494262e+06 7.92560098e+06
1.03845459e+07 8.60850001e+06 3.32130508e+06 -3.32078723e+06
-8.60853106e+06 -1.03850934e+07 -7.92608511e+06 -2.23485030e+06
4.36727105e+06 9.18801534e+06 1.02619538e+07 7.15144826e+06
1.12477815e+06 -5.36083671e+06 -9.66074764e+06 -1.00211059e+07
-6.29440755e+06 -2.26911460e+02]

e) Faça uma análise da sensibilidade do sistema linear do item (d)

In [4]: import numpy as np

def distance(p1, p2):

return np.linalg.norm(p1 - p2)

def matrix_U(p1, p2):

r = distance(p1, p2)
if r == 0:
return 0
 return 1 / (2 * np.pi) * np.log(1 / r)

def solve_system(U_matrix, b_vector):

U, s, Vt = np.linalg.svd(U_matrix)
S_inv = np.zeros((U_matrix.shape[1], U_matrix.shape[0]))
S_inv[:U_matrix.shape[1], :U_matrix.shape[1]] = np.linalg.inv(np.diag(s))
v = Vt.T @ S_inv @ U.T @ b_vector
return v

# Conjunto de pontos P1
t1_values = np.linspace(0, 2 * np.pi, 60)
P1 = np.array([(np.cos(t), np.sin(t)) for t in t1_values])

# Vetor b
b = np.sin(3 * t1_values)

# Valores de k com ligeiras variações

k_values = [1.05, 1.1, 1.2, 1.3, 1.5, 1.8, 2.0, 2.3, 2.6, 3, 3.5, 4, 5, 6.5, 8, 10, 20, 50, 100, 200]

# Armazenar as soluções para diferentes valores de k

solutions = {}

for k in k_values:
# Conjunto de pontos P2
t2_values = np.linspace(0, 2 * np.pi, 30)
P2 = np.array([k * np.array([np.cos(t), np.sin(t)]) for t in t2_values])

# Criar a matriz U para o valor atual de k

U_matrix = np.zeros((60, 30))

for i in range(60):
for j in range(30):
U_matrix[i, j] = matrix_U(P1[i], P2[j])

# Resolver o sistema linear para obter o vetor v

v = solve_system(U_matrix, b)

# Armazenar a solução para o valor atual de k

solutions[k] = v

# Comparar as soluções para diferentes valores de k

for k, v in solutions.items():
print(f"Solução para k = {k}:")
print(v)
print()

# Calcular a norma dos vetores solução

norms = {k: np.linalg.norm(v) for k, v in solutions.items()}

# Encontrar o valor de k mais sensível (maior norma)

most_sensitive_k = max(norms, key=norms.get)
print(f"Valor de k mais sensível: {most_sensitive_k}")
Solução para k = 1.05:
[ 5.09402741e+13 8.37686584e-01 1.43912866e+00 1.39180591e+00
7.83215313e-01 -1.56672386e-01 -1.02089234e+00 -1.47603088e+00
-1.34027069e+00 -6.44108534e-01 3.07173463e-01 1.15457372e+00
1.49353920e+00 1.27348020e+00 4.87718723e-01 -4.94712201e-01
-1.22846716e+00 -1.53300547e+00 -1.14684636e+00 -3.19974096e-01
6.13422835e-01 1.34680025e+00 1.46971644e+00 1.03688352e+00
1.55237659e-01 -7.70732599e-01 -1.39382201e+00 -1.43455220e+00
-8.74481501e-01 -5.09402741e+13]

Solução para k = 1.1:

[-2.81681738e+12 1.05158057e+00 1.66806695e+00 1.60320840e+00
8.93118504e-01 -1.85574868e-01 -1.18868387e+00 -1.70656222e+00
-1.52993998e+00 -7.26322245e-01 3.70914921e-01 1.31707750e+00
1.73150751e+00 1.43224508e+00 5.52410716e-01 -5.53255167e-01
-1.43160753e+00 -1.72950784e+00 -1.31842446e+00 -3.71773804e-01
7.27816404e-01 1.52883598e+00 1.70681756e+00 1.18917319e+00
1.86209393e-01 -8.91739413e-01 -1.60626286e+00 -1.66503845e+00
-1.04592041e+00 2.81681738e+12]

Solução para k = 1.2:

[ 7.39685386e+11 1.36127585e+00 2.16273223e+00 2.08757420e+00
1.15712954e+00 -2.42269981e-01 -1.54460391e+00 -2.21772893e+00
-1.98311198e+00 -9.44632584e-01 4.84066013e-01 1.71074308e+00
2.24427369e+00 1.85835539e+00 7.18003415e-01 -7.18405402e-01
-1.85855522e+00 -2.24387677e+00 -1.71117024e+00 -4.83330982e-01
9.42965300e-01 1.98527698e+00 2.21605254e+00 1.54534016e+00
2.42680402e-01 -1.15834196e+00 -2.08600271e+00 -2.16596716e+00
-1.35636361e+00 -7.39685386e+11]
Solução para k = 1.3:
[-1.57712080e+11 1.72842877e+00 2.75206792e+00 2.65331356e+00
1.47207460e+00 -3.08348203e-01 -1.96446132e+00 -2.81808808e+00
-2.52379549e+00 -1.19862877e+00 6.13418630e-01 2.17712481e+00
2.85177289e+00 2.36392092e+00 9.11741417e-01 -9.11451408e-01
-2.36454634e+00 -2.85119898e+00 -2.17717449e+00 -6.13775714e-01
1.19919412e+00 2.52318095e+00 2.81870942e+00 1.96379415e+00
3.08908169e-01 -1.47234619e+00 -2.65342389e+00 -2.75168394e+00
-1.72846185e+00 1.57712080e+11]

Solução para k = 1.5:

[ 3.38018353e+09 2.65511446e+00 4.22750754e+00 4.07576241e+00
2.26196461e+00 -4.74389441e-01 -3.01718992e+00 -4.32959848e+00
-3.87633531e+00 -1.84220584e+00 9.43161278e-01 3.34389581e+00
4.38098089e+00 3.63137767e+00 1.40091766e+00 -1.40090149e+00
-3.63141887e+00 -4.38093892e+00 -3.34392017e+00 -9.43160220e-01
1.84222805e+00 3.87631250e+00 4.32959469e+00 3.01722014e+00
4.74356543e-01 -2.26194469e+00 -4.07578047e+00 -4.22748035e+00
-2.65513765e+00 -3.38018352e+09]

Solução para k = 1.8:

[-1.86518909e+08 4.58806857e+00 7.30509551e+00 7.04294324e+00
3.90865323e+00 -8.19716106e-01 -5.21372522e+00 -7.48155225e+00
-6.69828362e+00 -3.18334018e+00 1.62976092e+00 5.77829037e+00
7.57028911e+00 6.27506844e+00 2.42075250e+00 -2.42075557e+00
-6.27505802e+00 -7.57030861e+00 -5.77826351e+00 -1.62978833e+00
3.18335953e+00 6.69827855e+00 7.48154092e+00 5.21374946e+00
8.19686738e-01 -3.90862596e+00 -7.04296573e+00 -7.30507669e+00
-4.58808515e+00 1.86518909e+08]

Solução para k = 2.0:

[-3.72023131e+06 6.29365937e+00 1.00206870e+01 9.66112815e+00
5.36163484e+00 -1.12441126e+00 -7.15190678e+00 -1.02627527e+01
-9.18831031e+00 -4.36674229e+00 2.23563844e+00 7.92630017e+00
1.03845012e+01 8.60776442e+00 3.32065610e+00 -3.32065613e+00
-8.60776439e+00 -1.03845011e+01 -7.92630068e+00 -2.23563751e+00
4.36674103e+00 9.18831180e+00 1.02627508e+01 7.15190936e+00
1.12440748e+00 -5.36162957e+00 -9.66113498e+00 -1.00206787e+01
-6.29366892e+00 3.72023131e+06]

Solução para k = 2.3:

[ 1.21867312e+05 9.57187621e+00 1.52402070e+01 1.46933719e+01
8.15437465e+00 -1.71008909e+00 -1.08771546e+01 -1.56083664e+01
-1.39742688e+01 -6.64127179e+00 3.40012917e+00 1.20549092e+01
1.57935310e+01 1.30913311e+01 5.05030537e+00 -5.05030532e+00
-1.30913312e+01 -1.57935308e+01 -1.20549095e+01 -3.40012875e+00
6.64127136e+00 1.39742692e+01 1.56083661e+01 1.08771547e+01
1.71008938e+00 -8.15437529e+00 -1.46933709e+01 -1.52402082e+01
-9.57187485e+00 -1.21867375e+05]

Solução para k = 2.6:

[ 1.14015703e+04 1.38271792e+01 2.20154421e+01 2.12255034e+01
1.17795093e+01 -2.47033153e+00 -1.57127373e+01 -2.25472707e+01
-2.01867144e+01 -9.59373613e+00 4.91170093e+00 1.74140781e+01
2.28147528e+01 1.89112548e+01 7.29548496e+00 -7.29548498e+00
-1.89112548e+01 -2.28147529e+01 -1.74140780e+01 -4.91170101e+00
9.59373622e+00 2.01867143e+01 2.25472708e+01 1.57127371e+01
2.47033180e+00 -1.17795097e+01 -2.12255030e+01 -2.20154427e+01
-1.38271785e+01 -1.14015938e+04]

Solução para k = 3:
[-288.5625 21.24111448 33.81980801 32.60631467 18.09551408
-3.79488814 -24.13768236 -34.63679491 -31.01054221 -14.73776022
7.54528465 26.75125795 35.04769709 29.05119946 11.20721963
-11.20721964 -29.05119945 -35.0476971 -26.75125793 -7.54528466
14.73776022 31.01054221 34.63679489 24.13768238 3.7948881
-18.09551402 -32.60631474 -33.81980791 -21.24111459 288.5625 ]

Solução para k = 3.5:

[ 2.0625 33.73010334 53.70460231 51.77762023 28.73500602
-6.0261417 -38.32974569 -55.00194738 -49.24359258 -23.40301728
11.98163245 42.48000698 55.65444474 46.13222896 17.79664948
-17.79664948 -46.13222897 -55.65444474 -42.48000697 -11.98163247
23.40301733 49.24359251 55.00194748 38.32974556 6.02614188
-28.73500624 -51.77761997 -53.70460262 -33.730103 -2.09375 ]

Solução para k = 4:
[ 24.5625 50.34931326 80.16546626 77.28904645 42.89306653
-8.99528687 -57.21525031 -82.1020294 -73.50647322 -34.93394744
17.88511642 63.41039207 83.07601976 68.86210546 26.56525849
-26.56525879 -68.86210449 -83.07602135 -63.41038986 -17.88511931
34.93395112 73.5064687 82.10203476 57.21524404 8.99529397
-42.89307438 -77.28903785 -80.16547543 -50.34930369 -24.5625 ]

Solução para k = 5:
[ -31.71875 98.33873844 156.57296658 150.95534754 83.77537203
-17.56880069 -111.74862933 -160.35545659 -143.56738079 -68.23033321
34.93185329 123.84841859 162.25787508 134.49625462 51.88533843
-51.88536406 -134.49618244 -162.25799131 -123.84826159 -34.93204689
68.23055983 143.56712246 160.35574794 111.74829996 17.56917095
-83.77578712 -150.95489025 -156.57346106 -98.33821774 31.765625 ]

Solução para k = 6.5:

[ 13.46875 216.05303955 343.9874115 331.65283203 184.05003357
-38.59366608 -245.51726532 -352.29496765 -315.42390442 -149.89539719
76.7384758 272.10181808 356.47376633 295.49493408 113.98562241
-113.98590088 -295.49414444 -356.4749794 -272.10021591 -76.74041367
149.89759064 315.42145157 352.29764557 245.51413727 38.597229
-184.0541687 -331.6481781 -343.99255371 -216.0475769 -13.40625 ]

Solução para k = 8:
[ -29.390625 402.20361328 641.83032227 617.8984375 343.46533203
-72.19262695 -457.57458496 -656.89282227 -588.03491211 -279.43933105
143.00823975 507.38861084 664.47442627 551.05731201 212.33398438
-212.30322266 -551.15203857 -664.30877686 -507.63562012 -142.66815186
278.99389648 588.59692383 656.20666504 458.3894043 71.25024414
-342.40405273 -619.06420898 -640.58251953 -403.50488281 30.078125 ]

Solução para k = 10:

[ -43.890625 120.3125 1907.25 572.5625 1278.875
-715.5625 -358.875 -1771.625 -711.5625 -925.5625
597.75 737.90625 1482.34375 962.390625 456.375
-383.4921875 -1180.21875 -1122.34375 -1236.0625 32.53125
171.0625 1581.875 796.25 1428.0625 -434.25
-61.1875 -1843.4375 -597.5 -1453.625 714.4375 ]

Solução para k = 20:

[ 12867.375 4377.40625 11480.15625 8755.3671875
5685.5625 -903.671875 -7811.0234375 -9341.359375
-10144.4375 -3606.640625 1838.15625 7919.875
10689.5625 8219.09375 3556.1640625 -3247.4375
-8989.875 -9854.1875 -8350.1875 -2150.65625
4719.28515625 8491.40625 11041.109375 6633.640625
1082.109375 -4597.65625 -11126.7421875 -8032.390625
-8532.3125 -10667.1875 ]

Solução para k = 50:

[ 344744.75 -6430.234375 175826.6796875 190424.921875
52762.2578125 -31942.734375 -74123.9375 -178228.2890625
-156763.921875 -50273.09375 37445.3125 106876.9375
169154.5 147264.09375 34992.515625 -48587.5625
-127957.76953125 -159720.36328125 -142574. -16355.671875
69944.890625 123442.65625 177630.640625 108847.40625
20350.75 -103075.4609375 -132195.78125 -127223.4375
-202864. -201391. ]

Solução para k = 100:

[ 540627.875 657480.15625 1150224.375 1355473.703125
610008.453125 -103954.1796875 -994934.4453125 -1166946.640625
-1199616.2265625 -525062.65429688 222909.1640625 1046698.07421875
1303996.6171875 1061139.88867188 346344.7421875 -296155.8984375
-1128087.359375 -1331836.671875 -965890.8046875 -244696.74023438
502894.0703125 1163174.25 1246030.921875 989444.203125
66603.6484375 -722853.4375 -1083211.46875 -1243166.3515625
-1034153.875 -222483.625 ]

Solução para k = 200:

[ 2200976.625 4430915.9375 10288094.34765625
9846689.9609375 5757759.47265625 -1537985.859375
-7653692.2734375 -9392267.79296875 -9006524.234375
-5727481.1484375 3276593.4765625 8258653.21484375
9514223.33398438 8486205.453125 4631116.44921875
-4662280.1328125 -7977561.796875 -10764243.37890625
-7325691.265625 -2600458.1875 4016179.9609375
9765650.8984375 10155401.859375 6949611.890625
1198068.4296875 -5531213.765625 -9218805.671875
-9831525.5703125 -8109538.4375 563128.625 ]

Valor de k mais sensível: 1.05

f) Calcule a solução numérica do problema (1) para os pontos y_1 = (0,0), y_2 = (0.7, 0.4), y_3 = (0, 0.7), y_4 = (-0.7, 0.4), y_5 = (-0.7, -0.4), y_6 = (0, -
0.7) e y_7 = (0.7, -0.4). Basta calcular:ϕ(yi )=∑(j=1)^30 u_ij^ (r)v_j , para i = 1, ..., 7. Os v_j são as componentes do vetor obtido no item (d). Observe que
a conta do u^ agora deve ser feita com os y_i.
In [5]: import numpy as np
 def distance(p1, p2):
return np.linalg.norm(p1 - p2)

def matrix_U(p1, p2):

r = distance(p1, p2)
if r == 0:
return 0
return 1 / (2 * np.pi) * np.log(1 / r)

def solve_system(U_matrix, b_vector):

U, s, Vt = np.linalg.svd(U_matrix)
S_inv = np.zeros((U_matrix.shape[1], U_matrix.shape[0]))
S_inv[:U_matrix.shape[1], :U_matrix.shape[1]] = np.linalg.inv(np.diag(s))
v = Vt.T @ S_inv @ U.T @ b_vector
return v

# Conjunto de pontos P1
t1_values = np.linspace(0, 2 * np.pi, 60)
P1 = np.array([(np.cos(t), np.sin(t)) for t in t1_values])

# Vetor b
b = np.sin(3 * t1_values)

# Valores de k para análise de sensibilidade

k_values = [1.05, 1.1, 1.2, 1.3, 1.5, 1.8, 2.0, 2.3, 2.6, 3, 3.5, 4, 5, 6.5, 8, 10, 20, 50, 100, 200]

# Armazenar as soluções para diferentes valores de k

solutions_list = []

for k in k_values:
# Conjunto de pontos P2
t2_values = np.linspace(0, 2 * np.pi, 30)
P2 = np.array([k * np.array([np.cos(t), np.sin(t)]) for t in t2_values])

# Criar a matriz U para o valor atual de k

U_matrix = np.zeros((60, 30))

for i in range(60):
for j in range(30):
U_matrix[i, j] = matrix_U(P1[i], P2[j])

# Resolver o sistema linear para obter o vetor v

v = solve_system(U_matrix, b)

# Armazenar a solução para o valor atual de k

solutions_list.append((k, v))

# Pontos y_i
y_points = np.array([(0, 0), (0.7, 0.4), (0, 0.7), (-0.7, 0.4), (-0.7, -0.4), (0, -0.7), (0.7, -0.4)])

# Calcular ϕ(y_i) para cada ponto y_i usando a matriz U e o vetor v

for k, v in solutions_list:
print(f"--- Resultados para k = {k} ---")
for i, y in enumerate(y_points):
u_star = np.zeros(30)
for j in range(30):
u_star[j] = matrix_U(y, k * np.array([np.cos(t2_values[j]), np.sin(t2_values[j])]))
phi_yi = np.sum(u_star * v)
print(f"ϕ(y_{i+1}) = {phi_yi:.15e}")
--- Resultados para k = 1.05 ---
ϕ(y_1) = 3.234863281250000e-03
ϕ(y_2) = 4.814453125000000e-01
ϕ(y_3) = -3.281250000000000e-01
ϕ(y_4) = 5.634765625000000e-01
ϕ(y_5) = -4.912109375000000e-01
ϕ(y_6) = 3.562011718750000e-01
ϕ(y_7) = -5.576171875000000e-01
--- Resultados para k = 1.1 ---
ϕ(y_1) = -3.913879394531250e-03
ϕ(y_2) = 5.469665527343750e-01
ϕ(y_3) = -3.537597656250000e-01
ϕ(y_4) = 4.992980957031250e-01
ϕ(y_5) = -5.493469238281250e-01
ϕ(y_6) = 3.320159912109375e-01
ϕ(y_7) = -5.003051757812500e-01
--- Resultados para k = 1.2 ---
ϕ(y_1) = 7.190704345703125e-03
ϕ(y_2) = 5.064468383789062e-01
ϕ(y_3) = -3.300552368164062e-01
ϕ(y_4) = 5.500946044921875e-01
ϕ(y_5) = -4.978942871093750e-01
ϕ(y_6) = 3.559799194335938e-01
ϕ(y_7) = -5.414810180664062e-01
--- Resultados para k = 1.3 ---
ϕ(y_1) = -1.049041748046875e-05
ϕ(y_2) = 5.239982604980469e-01
ϕ(y_3) = -3.430175781250000e-01
ϕ(y_4) = 5.239906311035156e-01
ϕ(y_5) = -5.240211486816406e-01
ϕ(y_6) = 3.429813385009766e-01
ϕ(y_7) = -5.239925384521484e-01
--- Resultados para k = 1.5 ---
ϕ(y_1) = -1.613238453865051e-02
ϕ(y_2) = 5.284391790628433e-01
ϕ(y_3) = -3.630522489547729e-01
ϕ(y_4) = 4.919821619987488e-01
ϕ(y_5) = -5.560178160667419e-01
ϕ(y_6) = 3.229479193687439e-01
ϕ(y_7) = -5.195608958601952e-01
--- Resultados para k = 1.8 ---
ϕ(y_1) = 5.847558379173279e-03
ϕ(y_2) = 5.255659380927682e-01
ϕ(y_3) = -3.364519067108631e-01
ϕ(y_4) = 5.332413949072361e-01
ϕ(y_5) = -5.147586092352867e-01
ϕ(y_6) = 3.495481088757515e-01
ϕ(y_7) = -5.224340567365289e-01
--- Resultados para k = 2.0 ---
ϕ(y_1) = 5.553010851144791e-07
ϕ(y_2) = 5.240002461359836e-01
ϕ(y_3) = -3.429993983008899e-01
ϕ(y_4) = 5.240008046384901e-01
ϕ(y_5) = -5.239991954294965e-01
ϕ(y_6) = 3.430006015114486e-01
ϕ(y_7) = -5.239997537864838e-01
--- Resultados para k = 2.3 ---
ϕ(y_1) = 8.285006681035156e-03
ϕ(y_2) = 5.289766784753738e-01
ϕ(y_3) = -3.342744104738813e-01
ϕ(y_4) = 5.350156167660316e-01
ϕ(y_5) = -5.129843832291954e-01
ϕ(y_6) = 3.517255895203562e-01
ϕ(y_7) = -5.190233215289481e-01
--- Resultados para k = 2.6 ---
ϕ(y_1) = 3.564190937140665e-03
ϕ(y_2) = 5.264750873982393e-01
ϕ(y_3) = -3.393052941282804e-01
ϕ(y_4) = 5.284807002735761e-01
ϕ(y_5) = -5.195192997289269e-01
ϕ(y_6) = 3.466947058623191e-01
ϕ(y_7) = -5.215249126090384e-01
--- Resultados para k = 3 ---
ϕ(y_1) = 9.462617356348346e-09
ϕ(y_2) = 5.240000072561131e-01
ϕ(y_3) = -3.429999902650920e-01
ϕ(y_4) = 5.240000113672494e-01
ϕ(y_5) = -5.239999886575362e-01
ϕ(y_6) = 3.430000096871524e-01
ϕ(y_7) = -5.239999927767158e-01
--- Resultados para k = 3.5 ---
ϕ(y_1) = 6.230694333023923e-03
ϕ(y_2) = 5.291711149481683e-01
ϕ(y_3) = -3.366717724130708e-01
ϕ(y_4) = 5.311599364245789e-01
ϕ(y_5) = -5.168400633266343e-01
ϕ(y_6) = 3.493282280095920e-01
ϕ(y_7) = -5.188288848024669e-01
--- Resultados para k = 4 ---
ϕ(y_1) = -1.082484012648877e-06
ϕ(y_2) = 5.239990609441687e-01
ϕ(y_3) = -3.430010935679508e-01
ϕ(y_4) = 5.239987907744057e-01
ϕ(y_5) = -5.240012103816865e-01
ϕ(y_6) = 3.429989050547029e-01
ϕ(y_7) = -5.240009397326162e-01
--- Resultados para k = 5 ---
ϕ(y_1) = -1.207431610504983e-02
ϕ(y_2) = 5.130249184806095e-01
ϕ(y_3) = -3.551471391292313e-01
ϕ(y_4) = 5.109242009666701e-01
ϕ(y_5) = -5.370758033682641e-01
ϕ(y_6) = 3.308528786156337e-01
ϕ(y_7) = -5.349750580207173e-01
--- Resultados para k = 6.5 ---
ϕ(y_1) = -1.941011122727287e-02
ϕ(y_2) = 5.057492771426952e-01
ϕ(y_3) = -3.624690319000203e-01
ϕ(y_4) = 5.035124058487908e-01
ϕ(y_5) = -5.444887724942236e-01
ϕ(y_6) = 3.235288118412560e-01
ϕ(y_7) = -5.422520087255864e-01
--- Resultados para k = 8 ---
ϕ(y_1) = -8.988907659862377e-03
ϕ(y_2) = 5.153986308654144e-01
ϕ(y_3) = -3.520120008411673e-01
ϕ(y_4) = 5.146387273201505e-01
ϕ(y_5) = -5.333522784630098e-01
ϕ(y_6) = 3.340010468623724e-01
ϕ(y_7) = -5.325961258578538e-01
--- Resultados para k = 10 ---
ϕ(y_1) = -2.891659979951555e-01
ϕ(y_2) = 2.390439047047721e-01
ϕ(y_3) = -6.294845196009078e-01
ϕ(y_4) = 2.330286102668424e-01
ϕ(y_5) = -8.174378571349052e-01
ϕ(y_6) = 5.169537172020000e-02
ϕ(y_7) = -8.120297334488100e-01
--- Resultados para k = 20 ---
ϕ(y_1) = -2.402552435132748e-01
ϕ(y_2) = 2.823262049732875e-01
ϕ(y_3) = -5.828742217927356e-01
ϕ(y_4) = 2.856300267576444e-01
ϕ(y_5) = -7.628160915946864e-01
ϕ(y_6) = 1.023376144112262e-01
ϕ(y_7) = -7.661259371761844e-01
--- Resultados para k = 50 ---
ϕ(y_1) = -6.566672095796093e-01
ϕ(y_2) = -1.302825181046501e-01
ϕ(y_3) = -9.995511201850604e-01
ϕ(y_4) = -1.348904522892553e-01
ϕ(y_5) = -1.183030038169818e+00
ϕ(y_6) = -3.138064799713902e-01
ϕ(y_7) = -1.178434607660165e+00
--- Resultados para k = 100 ---
ϕ(y_1) = 1.732132959878072e-01
ϕ(y_2) = 6.971218862163369e-01
ϕ(y_3) = -1.697594805445988e-01
ϕ(y_4) = 6.973351784108672e-01
ϕ(y_5) = -3.506971432070713e-01
ϕ(y_6) = 5.161865929549094e-01
ϕ(y_7) = -3.509074382891413e-01
--- Resultados para k = 200 ---
ϕ(y_1) = -3.541005430743098e-01
ϕ(y_2) = 1.700044833123684e-01
ϕ(y_3) = -6.970835900865495e-01
ϕ(y_4) = 1.698138338397257e-01
ϕ(y_5) = -8.782056414638646e-01
ϕ(y_6) = -1.111752865836024e-02
ϕ(y_7) = -8.780148327350616e-01
(g) Compare a resposta nos pontos do item (f) com a solução exata.

(h) Para qual r a solução foi a melhor dentro da sua comparação?

In [6]: import numpy as np
import matplotlib.pyplot as plt

def distance(p1, p2):

return np.linalg.norm(p1 - p2)

# Solução exata para ϕ(r, θ) = r^3 * sen(3θ)

def phi_exata(r, theta):
return r**3 * np.sin(3 * theta)

def matrix_U(p1, p2):

r = distance(p1, p2)
if r == 0:
return 0
return 1 / (2 * np.pi) * np.log(1 / r)

def solve_system(U_matrix, b_vector):

U, s, Vt = np.linalg.svd(U_matrix)
S_inv = np.zeros((U_matrix.shape[1], U_matrix.shape[0]))
S_inv[:U_matrix.shape[1], :U_matrix.shape[1]] = np.linalg.inv(np.diag(s))
v = Vt.T @ S_inv @ U.T @ b_vector
return v

# Conjunto de pontos P1
t1_values = np.linspace(0, 2 * np.pi, 60)
P1 = np.array([(np.cos(t), np.sin(t)) for t in t1_values])

# Vetor b
b = np.sin(3 * t1_values)

# Valores de k para análise de sensibilidade

k_values = [1.05, 1.1, 1.2, 1.3, 1.5, 1.8, 2.0, 2.3, 2.6, 3, 3.5, 4, 5, 6.5, 8, 10, 20, 50, 100, 200]

# Armazenar as soluções para diferentes valores de k

solutions_list = []

for k in k_values:
# Conjunto de pontos P2
t2_values = np.linspace(0, 2 * np.pi, 30)
P2 = np.array([k * np.array([np.cos(t), np.sin(t)]) for t in t2_values])

# Criar a matriz U para o valor atual de k

U_matrix = np.zeros((60, 30))

for i in range(60):
for j in range(30):
U_matrix[i, j] = matrix_U(P1[i], P2[j])

# Resolver o sistema linear para obter o vetor v

v = solve_system(U_matrix, b)

# Armazenar a solução para o valor atual de k

solutions_list.append({k: v})

# Pontos y_i
y_points = np.array([(0, 0), (0.7, 0.4), (0, 0.7), (-0.7, 0.4), (-0.7, -0.4), (0, -0.7), (0.7, -0.4)])

# Valores de theta_i para cada ponto y_i

theta_values = np.arctan2(y_points[:, 1], y_points[:, 0])

# Cálculo da solução exata para cada ponto y_i

phi_exata_values = phi_exata(np.linalg.norm(y_points, axis=1), theta_values)

# Calcular o erro para cada valor de k

error_values = []

for solutions_dict in solutions_list:

k, v = next(iter(solutions_dict.items()))
phi_values_calculated = []
for y in y_points:
u_star = np.zeros(30)
for j in range(30):
u_star[j] = matrix_U(y, k * np.array([np.cos(t2_values[j]), np.sin(t2_values[j])]))
phi_yi = np.sum(u_star * v)
phi_values_calculated.append(phi_yi)
error_values.append(np.abs(phi_values_calculated - phi_exata_values))

# Encontrar o valor de k que resulta no menor erro absoluto médio

mean_error_values = np.mean(error_values, axis=1)
best_k_index = np.argmin(mean_error_values)
best_k = k_values[best_k_index]

# Imprimir os resultados
for i, (phi_calculated, phi_exact, error, r) in enumerate(zip(phi_values_calculated, phi_exata_values, error_values[best_k_index], np.linalg.norm(y_point
print(f"Para o ponto y_{i+1}:")
print(f"Valor calculado de ϕ(y_{i+1}) = {phi_calculated}")
print(f"Valor exato de ϕ(y_{i+1}) = {phi_exact}")
print(f"Erro absoluto = {error}")
print(f"Valor de r = {r}")
print()

print(f"O valor de k que resulta no menor erro absoluto médio é: {best_k}")

# Valores de k com ligeiras variações

k_values = [1.05, 1.1, 1.2, 1.3, 1.5, 1.8, 2.0, 2.3, 2.6, 3, 3.5, 4, 5, 6.5, 8, 10, 20, 50, 100, 200]

# Lista para armazenar os erros para cada valor de k em cada ponto y_i
errors_per_k = []

for k in k_values:
# Conjunto de pontos P2
t2_values = np.linspace(0, 2 * np.pi, 30)
P2 = np.array([k * np.array([np.cos(t), np.sin(t)]) for t in t2_values])

# Criar a matriz U para o valor atual de k

U_matrix = np.zeros((60, 30))
 for i in range(60):
for j in range(30):
U_matrix[i, j] = matrix_U(P1[i], P2[j])

# Resolver o sistema linear para obter o vetor v

v = solve_system(U_matrix, b)

# Calcular ϕ(y_i) para cada ponto y_i usando a matriz U e o vetor v

phi_values = []
for y in y_points:
u_star = np.zeros(30)
for j in range(30):
u_star[j] = matrix_U(y, k * np.array([np.cos(t2_values[j]), np.sin(t2_values[j])]))
phi_yi = np.sum(u_star * v)
phi_values.append(phi_yi)

# Cálculo do erro absoluto para cada ponto y_i

errors = np.abs(phi_values - phi_exata_values)
errors_per_k.append(errors)

# Encontrar o valor de k que resulta no menor erro absoluto médio

mean_errors_per_k = np.mean(errors_per_k, axis=1)
best_k_index = np.argmin(mean_errors_per_k)
best_k = k_values[best_k_index]
Para o ponto y_1:
Valor calculado de ϕ(y_1) = -0.3541005430743098
Valor exato de ϕ(y_1) = 0.0
Erro absoluto = 9.462617356348346e-09
Valor de r = 0.0

Para o ponto y_2:

Valor calculado de ϕ(y_2) = 0.1700044833123684
Valor exato de ϕ(y_2) = 0.5239999999999999
Erro absoluto = 7.2561131547743685e-09
Valor de r = 0.8062257748298549

Para o ponto y_3:

Valor calculado de ϕ(y_3) = -0.6970835900865495
Valor exato de ϕ(y_3) = -0.3429999999999999
Erro absoluto = 9.734907935321502e-09
Valor de r = 0.7

Para o ponto y_4:

Valor calculado de ϕ(y_4) = 0.1698138338397257
Valor exato de ϕ(y_4) = 0.5239999999999999
Erro absoluto = 1.136724947503609e-08
Valor de r = 0.8062257748298549

Para o ponto y_5:

Valor calculado de ϕ(y_5) = -0.8782056414638646
Valor exato de ϕ(y_5) = -0.5239999999999999
Erro absoluto = 1.1342463746011333e-08
Valor de r = 0.8062257748298549

Para o ponto y_6:

Valor calculado de ϕ(y_6) = -0.011117528658360243
Valor exato de ϕ(y_6) = 0.3429999999999999
Erro absoluto = 9.687152469073368e-09
Valor de r = 0.7

Para o ponto y_7:

Valor calculado de ϕ(y_7) = -0.8780148327350616
Valor exato de ϕ(y_7) = -0.5239999999999999
Erro absoluto = 7.2232840819808075e-09
Valor de r = 0.8062257748298549

O valor de k que resulta no menor erro absoluto médio é: 3

===============================================================================================================

Solução com a Fatoração QR através de transformações de Householder

b) Faça um gráfico de uma função que retorna o número de condicionamento e o posto da matriz U, sendo o eixo-x a variação de r_ij e o eixo-y o valor
do número de condicionamento da matriz U. Escolha valores para K no intervalo (1, 200]
In [7]: import numpy as np

def distance(p1, p2):

return np.linalg.norm(p1 - p2)

def matrix_U(p1, p2):

 r = distance(p1, p2)
if r == 0:
return 0
return 1 / (2 * np.pi) * np.log(1 / r)

def matrix_rank(matrix):
return np.linalg.matrix_rank(matrix)

def condition_number(matrix):
return np.linalg.cond(matrix)

import matplotlib.pyplot as plt

# Valores de k fornecidos
k_values = [1.05, 1.1, 1.2, 1.3, 1.5, 1.8, 2.0, 2.3, 2.6, 3, 3.5, 4, 5, 6.5, 8, 10, 20, 50, 100, 200]

# Matriz retangular U, posto da matriz e número de condicionamento para cada valor de k

matrix_rank_values = []
num_conditioning = []

# Conjunto de pontos P1
t1_values = np.linspace(0, 2 * np.pi, 60)
P1 = np.array([(np.cos(t), np.sin(t)) for t in t1_values])

# Conjunto de pontos P2
t2_values = np.linspace(0, 2 * np.pi, 30)

for k in k_values:
U_matrix = np.zeros((60, 30))

for i in range(60):
for j in range(30):
P2_point = k * np.array([np.cos(t2_values[j]), np.sin(t2_values[j])])
U_matrix[i, j] = matrix_U(P1[i], P2_point)

rank = matrix_rank(U_matrix)
matrix_rank_values.append(rank)

num_cond = condition_number(U_matrix)
num_conditioning.append(num_cond)

# Plotagem do gráfico
plt.figure(figsize=(10, 6))
plt.plot(k_values, matrix_rank_values, marker='o', label='Posto da Matriz U')
plt.plot(k_values, num_conditioning, marker='o', label='Número de Condicionamento')
plt.xlabel('Valores de k')
plt.ylabel('Posto / Número de Condicionamento')
plt.title('Posto e Número de Condicionamento da Matriz U para diferentes valores de k')
plt.legend()
plt.grid(True)
plt.show()
c) Monte um vetor com 60 posições da forma: b = b_i = sen(3t_1)
In [8]: import numpy as np

# Sequência de ângulos t1
t1_values = np.linspace(0, 2 * np.pi, 60)

# Vetor b
b = np.sin(3 * t1_values)

print(b)
[ 0.00000000e+00 3.14076712e-01 5.96367359e-01 8.18302776e-01
9.57422038e-01 9.99645611e-01 9.40700267e-01 7.86551556e-01
5.52800065e-01 2.63102564e-01 -5.32221748e-02 -3.64160575e-01
-6.38244184e-01 -8.47734428e-01 -9.71429893e-01 -9.96812007e-01
-9.21311978e-01 -7.52570770e-01 -5.07665800e-01 -2.11382624e-01
1.06293486e-01 4.13212186e-01 6.78311836e-01 8.74763085e-01
9.82684125e-01 9.91152831e-01 8.99312130e-01 7.16456740e-01
4.61092501e-01 1.59063496e-01 -1.59063496e-01 -4.61092501e-01
-7.16456740e-01 -8.99312130e-01 -9.91152831e-01 -9.82684125e-01
-8.74763085e-01 -6.78311836e-01 -4.13212186e-01 -1.06293486e-01
2.11382624e-01 5.07665800e-01 7.52570770e-01 9.21311978e-01
9.96812007e-01 9.71429893e-01 8.47734428e-01 6.38244184e-01
3.64160575e-01 5.32221748e-02 -2.63102564e-01 -5.52800065e-01
-7.86551556e-01 -9.40700267e-01 -9.99645611e-01 -9.57422038e-01
-8.18302776e-01 -5.96367359e-01 -3.14076712e-01 -7.34788079e-16]
d) Resolva o sistema linear, 60 x 30. Uv = b com a Fatoração QR através de transformações de Householder;
In [9]: import numpy as np

# Função para calcular a matriz U para um determinado valor de k

def calculate_U(k):
t1_values = np.linspace(0, 2 * np.pi, 60)
t2_values = np.linspace(0, 2 * np.pi, 30)

P1 = np.array([(np.cos(t), np.sin(t)) for t in t1_values])

P2 = np.array([k * np.array([np.cos(t), np.sin(t)]) for t in t2_values])

U_matrix = np.zeros((60, 30))

for i in range(60):
for j in range(30):
U_matrix[i, j] = 1 / (2 * np.pi) * np.log(1 / np.linalg.norm(P1[i] - P2[j]))

return U_matrix

# Vetor b
t1_values = np.linspace(0, 2 * np.pi, 60)
b = np.sin(3 * t1_values)

# Valores de k fornecidos
 k_values = [1.05, 1.1, 1.2, 1.3, 1.5, 1.8, 2.0, 2.3, 2.6, 3, 3.5, 4, 5, 6.5, 8, 10, 20, 50, 100, 200]

# Solução aproximada para cada valor de k

solutions = []

for k in k_values:
U_matrix_k = calculate_U(k)

# Fatoração QR de U_matrix_k
Q, R = np.linalg.qr(U_matrix_k)

# Resolvendo o sistema triangular superior Rv = QTb usando substituição para frente

Qt_b = np.dot(Q.T, b)
v_approx = np.linalg.solve(R, Qt_b)

solutions.append(v_approx)

# Exibindo as soluções aproximadas para cada valor de k

for i, k in enumerate(k_values):
print(f"Resultados para k = {k}:")
print(f"Vetor v:",solutions[i])
print()
Resultados para k = 1.05:
Vetor v: [ 5.09402741e+13 8.51684812e-01 1.44189672e+00 1.38296125e+00
7.76412743e-01 -1.58322266e-01 -1.02357184e+00 -1.48127104e+00
-1.33449144e+00 -6.43343729e-01 3.20094792e-01 1.15010798e+00
1.51094108e+00 1.25547147e+00 4.84712011e-01 -4.89216521e-01
-1.25058624e+00 -1.51554869e+00 -1.14921087e+00 -3.19887411e-01
6.36617518e-01 1.33468981e+00 1.47985557e+00 1.02357325e+00
1.52978475e-01 -7.72966174e-01 -1.39036813e+00 -1.43383064e+00
-8.74097610e-01 -5.09402741e+13]

Resultados para k = 1.1:

Vetor v: [-2.81681738e+12 1.04968495e+00 1.66492410e+00 1.60611298e+00
8.91886421e-01 -1.86110590e-01 -1.18824615e+00 -1.70757319e+00
-1.52822592e+00 -7.27859499e-01 3.71417274e-01 1.31866802e+00
1.72891472e+00 1.43312379e+00 5.52809964e-01 -5.52882108e-01
-1.43304552e+00 -1.72872974e+00 -1.31884598e+00 -3.71296278e-01
7.27465235e-01 1.52927618e+00 1.70692245e+00 1.18859969e+00
1.86118498e-01 -8.91774541e-01 -1.60659439e+00 -1.66496194e+00
-1.04667299e+00 2.81681738e+12]

Resultados para k = 1.2:

Vetor v: [ 7.39685386e+11 1.36040178e+00 2.16407675e+00 2.08687085e+00
1.15806510e+00 -2.42901698e-01 -1.54470182e+00 -2.21680006e+00
-1.98461276e+00 -9.43296363e-01 4.83117262e-01 1.71178018e+00
2.24321070e+00 1.85930158e+00 7.17243544e-01 -7.17262810e-01
-1.85932857e+00 -2.24296251e+00 -1.71227019e+00 -4.82544059e-01
9.42834595e-01 1.98499700e+00 2.21647124e+00 1.54496760e+00
2.43047086e-01 -1.15870834e+00 -2.08583521e+00 -2.16588660e+00
-1.35664344e+00 -7.39685386e+11]

Resultados para k = 1.3:

Vetor v: [-1.57712080e+11 1.72879293e+00 2.75178291e+00 2.65329648e+00
1.47234076e+00 -3.08743727e-01 -1.96407442e+00 -2.81844235e+00
-2.52331818e+00 -1.19924068e+00 6.14014172e-01 2.17662052e+00
2.85208387e+00 2.36367690e+00 9.12075415e-01 -9.11982430e-01
-2.36391317e+00 -2.85181559e+00 -2.17681131e+00 -6.13876722e-01
1.19918091e+00 2.52327859e+00 2.81853485e+00 1.96395875e+00
3.08869812e-01 -1.47241608e+00 -2.65333151e+00 -2.75176469e+00
-1.72840899e+00 1.57712080e+11]

Resultados para k = 1.5:

Vetor v: [ 1.13863857e+10 2.65508411e+00 4.22754218e+00 4.07571017e+00
2.26203789e+00 -4.74468197e-01 -3.01713242e+00 -4.32961805e+00
-3.87634704e+00 -1.84217436e+00 9.43129005e-01 3.34390006e+00
4.38101092e+00 3.63133708e+00 1.40094228e+00 -1.40089847e+00
-3.63145045e+00 -4.38088950e+00 -3.34396327e+00 -9.43141328e-01
1.84222968e+00 3.87631196e+00 4.32957548e+00 3.01725225e+00
4.74336357e-01 -2.26195029e+00 -4.07575985e+00 -4.22749963e+00
-2.65512692e+00 -1.13863857e+10]

Resultados para k = 1.8:

Vetor v: [-1.40758365e+09 4.58802946e+00 7.30514231e+00 7.04287957e+00
3.90873683e+00 -8.19810131e-01 -5.21363424e+00 -7.48163131e+00
-6.69821847e+00 -3.18339250e+00 1.62980180e+00 5.77826019e+00
7.57030468e+00 6.27507225e+00 2.42073440e+00 -2.42073865e+00
-6.27505771e+00 -7.57032948e+00 -5.77823672e+00 -1.62980151e+00
3.18334906e+00 6.69831668e+00 7.48147303e+00 5.21383480e+00
8.19611869e-01 -3.90858069e+00 -7.04298675e+00 -7.30506028e+00
-4.58811210e+00 1.40758365e+09]
Resultados para k = 2.0:
Vetor v: [-1.75889094e+07 6.29364423e+00 1.00206976e+01 9.66112167e+00
5.36163763e+00 -1.12441060e+00 -7.15191025e+00 -1.02627476e+01
-9.18831592e+00 -4.36673667e+00 2.23563282e+00 7.92630585e+00
1.03844956e+01 8.60776987e+00 3.32065092e+00 -3.32065106e+00
-8.60776981e+00 -1.03844947e+01 -7.92630858e+00 -2.23562797e+00
4.36673017e+00 9.18832349e+00 1.02627385e+01 7.15192259e+00
1.12439306e+00 -5.36161370e+00 -9.66115251e+00 -1.00206595e+01
-6.29368891e+00 1.75889094e+07]

Resultados para k = 2.3:

Vetor v: [ 8.96295970e+05 9.57188016e+00 1.52402039e+01 1.46933739e+01
8.15437372e+00 -1.71008919e+00 -1.08771537e+01 -1.56083679e+01
-1.39742670e+01 -6.64127358e+00 3.40013079e+00 1.20549078e+01
1.57935320e+01 1.30913304e+01 5.05030584e+00 -5.05030567e+00
-1.30913310e+01 -1.57935307e+01 -1.20549098e+01 -3.40012824e+00
6.64127084e+00 1.39742694e+01 1.56083665e+01 1.08771534e+01
1.71009157e+00 -8.15437840e+00 -1.46933671e+01 -1.52402126e+01
-9.57187025e+00 -8.96295970e+05]

Resultados para k = 2.6:

Vetor v: [ 5.96924021e+04 1.38271805e+01 2.20154410e+01 2.12255044e+01
1.17795087e+01 -2.47033115e+00 -1.57127374e+01 -2.25472707e+01
-2.01867142e+01 -9.59373639e+00 4.91170117e+00 1.74140779e+01
2.28147529e+01 1.89112547e+01 7.29548511e+00 -7.29548519e+00
-1.89112545e+01 -2.28147533e+01 -1.74140775e+01 -4.91170155e+00
9.59373680e+00 2.01867137e+01 2.25472714e+01 1.57127364e+01
2.47033252e+00 -1.17795105e+01 -2.12255020e+01 -2.20154439e+01
-1.38271772e+01 -5.96924021e+04]

Resultados para k = 3:
Vetor v: [-2432.07944394 21.24111422 33.8198082 32.60631456
18.0955141 -3.79488809 -24.13768247 -34.63679475
-31.01054241 -14.73775998 7.54528438 26.75125824
35.04769679 29.05119975 11.20721937 -11.20721941
-29.05119965 -35.04769693 -26.75125808 -7.54528451
14.73776007 31.01054238 34.63679471 24.13768258
3.79488788 -18.09551377 -32.60631503 -33.8198076
-21.24111492 2432.07944431]

Resultados para k = 3.5:

Vetor v: [ 10.12795706 33.73010329 53.70460236 51.7776202 28.73500604
-6.02614172 -38.32974567 -55.0019474 -49.24359257 -23.40301729
11.98163245 42.48000698 55.65444473 46.13222898 17.79664947
-17.79664946 -46.13222899 -55.65444472 -42.48000699 -11.98163244
23.4030173 49.24359254 55.00194745 38.32974558 6.02614184
-28.7350062 -51.77762001 -53.70460258 -33.73010305 -10.12795718]

Resultados para k = 4:
Vetor v: [ 76.06189383 50.34931111 80.16546826 77.28904462 42.89306811
-8.99528823 -57.21524913 -82.10203039 -73.50647241 -34.93394807
17.88511691 63.4103917 83.07602002 68.86210525 26.56525855
-26.56525876 -68.86210463 -83.07602104 -63.41039026 -17.88511878
34.93395043 73.50646952 82.10203382 57.21524519 8.99529261
-42.89307284 -77.28903962 -80.16547344 -50.34930582 -76.0618965 ]

Resultados para k = 5:
Vetor v: [-744.5449948 98.33852192 156.57317982 150.9551416 83.77556899
-17.56898451 -111.74846055 -160.35560551 -143.56725221 -68.23043995
34.93193637 123.8483578 162.2579144 134.49623298 51.88534435
-51.88535429 -134.49620579 -162.25795185 -123.84831919 -34.93196785
68.23045947 143.56724459 160.35560607 111.74845883 17.56899673
-83.77559961 -150.95508833 -156.57325498 -98.33843089 744.54494639]

Resultados para k = 6.5:

Vetor v: [ 306.43778249 216.05340862 343.98712676 331.65307722 184.04980283
-38.59345636 -245.5174064 -352.29487325 -315.42388194 -149.89555163
76.73879331 272.10131567 356.47448265 295.49397416 113.98680268
-113.98728428 -295.49256309 -356.47672979 -272.09836407 -76.74230909
149.89951209 315.4195547 352.29953675 245.51240099 38.59882203
-184.05553397 -331.64700668 -343.9934734 -216.0468818 -306.44107719]

Resultados para k = 8:
Vetor v: [-1214.43437347 402.2228963 641.8154133 617.91153044
343.44967843 -72.17289463 -457.59975705 -656.85985093
-588.07518 -279.3932681 142.95920507 507.43813679
664.42802543 551.09657738 212.3051049 -212.2867667
-551.15499695 -664.31983067 -507.61063813 -142.70510944
279.03985542 588.54385595 656.26364009 458.33054389
71.30701755 -342.45581932 -619.01785231 -640.62071354
-403.47424164 1215.06981194]
Resultados para k = 10:
Vetor v: [-3928.00303421 757.3640925 1280.5832034 1181.43105984
694.30294583 -162.32793504 -874.64561916 -1299.72264427
-1134.09012259 -557.94433093 289.32158747 983.03823097
1303.7869318 1072.23256742 416.79915846 -414.66777746
-1078.68679752 -1292.84351036 -998.70578752 -268.67734265
532.10556731 1165.2614733 1263.19561486 916.40832931
115.61772893 -643.12452612 -1236.39148791 -1222.73328406
-817.03104254 3958.14675094]

Resultados para k = 20:

Vetor v: [ 10949.53949015 9965.22338565 4375.73770137 16358.10441322
-1390.45833106 4849.11597682 -11824.12490814 -7067.86422699
-11026.00056514 -3559.95172602 2017.67562341 8037.15710563
9927.73542266 9763.23987239 1286.30842939 -451.17826139
-12047.54730064 -6780.4451516 -11295.78554688 676.14320354
1845.50510656 11671.39654165 7324.56871983 10941.05565278
-3579.15786208 -138.00421737 -14602.54888366 -6353.82027105
-7810.54135493 -12061.07803811]

Resultados para k = 50:

Vetor v: [ 84632.5364527 62756.36513771 119911.86335453 222972.72443061
75261.81127623 -133980.82572547 75012.1639139 -288096.61070468
-153528.30194208 34653.69265622 -51900.50251902 141050.81829563
169834.34922224 171059.32886203 -28504.54707006 5184.59571784
-125126.05646937 -194457.71439951 -152060.52208958 72942.36211717
-38420.85092719 145027.02808804 287152.06489592 -68627.61472053
158250.18089256 -150197.48182684 -138410.24909811 -135998.62036204
-151014.73343475 -15377.25402412]

Resultados para k = 100:

Vetor v: [-1047642.42433127 1198754.37723802 584879.73364253 1618727.09219034
460409.75418747 268745.27850576 -1567576.68607678 -675560.38491809
-1515458.26928068 -243610.86961855 -71699.90648465 1230886.09280524
1245992.81483573 1185378.53731372 26306.47981981 86092.45185094
-1328198.27924379 -1437144.40733872 -603020.15179506 -668215.85645359
661944.69573037 1527405.38040133 495568.31539767 1574010.38353347
107288.69042517 -1235536.41627195 -712076.32468896 -1118337.35989568
-1337567.16472755 1289254.42324774]

Resultados para k = 200:

Vetor v: [ -1530667.34465275 2626698.29492198 18065487.27119772
4888817.47732183 6665768.29796466 -3664328.57649286
-6208588.15002503 -4407422.76560693 -15837619.7088785
-6379547.2631297 8744281.72893899 7545645.93484305
7173775.14258692 3980355.90206967 16416481.10826211
-15011409.58571111 -2694177.41155057 -13531190.65030401
-6829804.58961376 -172209.03779668 1858486.28000782
9592732.94746244 10010712.17837579 7997067.67427783
2858205.0394027 -9584690.1866314 -7323408.77124175
-10520038.65885304 -3641040.83419864 -1088371.7429468 ]

e) Faça uma análise da sensibilidade do sistema linear do item (d)

In [10]: import numpy as np

# Função para calcular a matriz U para um determinado valor de k

def calculate_U(k):
t1_values = np.linspace(0, 2 * np.pi, 60)
t2_values = np.linspace(0, 2 * np.pi, 30)

P1 = np.array([(np.cos(t), np.sin(t)) for t in t1_values])

P2 = np.array([k * np.array([np.cos(t), np.sin(t)]) for t in t2_values])

U_matrix = np.zeros((60, 30))

for i in range(60):
for j in range(30):
U_matrix[i, j] = 1 / (2 * np.pi) * np.log(1 / np.linalg.norm(P1[i] - P2[j]))

return U_matrix

# Vetor b
t1_values = np.linspace(0, 2 * np.pi, 60)
b = np.sin(3 * t1_values)

# Valores de k fornecidos
k_values = [1.05, 1.1, 1.2, 1.3, 1.5, 1.8, 2.0, 2.3, 2.6, 3, 3.5, 4, 5, 6.5, 8, 10, 20, 50, 100, 200]

# Soluções aproximadas e seus desvios para cada valor de k

solutions = []
sensitivity = []

for k in k_values:
 U_matrix_k = calculate_U(k)

# Fatoração QR de U_matrix_k
Q, R = np.linalg.qr(U_matrix_k)

# Resolvendo o sistema triangular superior Rv = QTb usando substituição para frente

Qt_b = np.dot(Q.T, b)
v_approx = np.linalg.solve(R, Qt_b)

# Calculando o desvio da solução aproximada

deviation = np.linalg.norm(U_matrix_k @ v_approx - b)
sensitivity.append(deviation)

solutions.append(v_approx)

# Exibindo as soluções aproximadas e seus desvios para cada valor de k

for i, k in enumerate(k_values):
print(f"--- Resultados para k = {k} ---")
print(f"Solução aproximada: {solutions[i]}")
print(f"Desvio da solução: {sensitivity[i]}")
print()
--- Resultados para k = 1.05 ---
Solução aproximada: [ 5.09402741e+13 8.51684812e-01 1.44189672e+00 1.38296125e+00
7.76412743e-01 -1.58322266e-01 -1.02357184e+00 -1.48127104e+00
-1.33449144e+00 -6.43343729e-01 3.20094792e-01 1.15010798e+00
1.51094108e+00 1.25547147e+00 4.84712011e-01 -4.89216521e-01
-1.25058624e+00 -1.51554869e+00 -1.14921087e+00 -3.19887411e-01
6.36617518e-01 1.33468981e+00 1.47985557e+00 1.02357325e+00
1.52978475e-01 -7.72966174e-01 -1.39036813e+00 -1.43383064e+00
-8.74097610e-01 -5.09402741e+13]
Desvio da solução: 0.23524621462685683

--- Resultados para k = 1.1 ---

Solução aproximada: [-2.81681738e+12 1.04968495e+00 1.66492410e+00 1.60611298e+00
8.91886421e-01 -1.86110590e-01 -1.18824615e+00 -1.70757319e+00
-1.52822592e+00 -7.27859499e-01 3.71417274e-01 1.31866802e+00
1.72891472e+00 1.43312379e+00 5.52809964e-01 -5.52882108e-01
-1.43304552e+00 -1.72872974e+00 -1.31884598e+00 -3.71296278e-01
7.27465235e-01 1.52927618e+00 1.70692245e+00 1.18859969e+00
1.86118498e-01 -8.91774541e-01 -1.60659439e+00 -1.66496194e+00
-1.04667299e+00 2.81681738e+12]
Desvio da solução: 0.07698865637605294

--- Resultados para k = 1.2 ---

Solução aproximada: [ 7.39685386e+11 1.36040178e+00 2.16407675e+00 2.08687085e+00
1.15806510e+00 -2.42901698e-01 -1.54470182e+00 -2.21680006e+00
-1.98461276e+00 -9.43296363e-01 4.83117262e-01 1.71178018e+00
2.24321070e+00 1.85930158e+00 7.17243544e-01 -7.17262810e-01
-1.85932857e+00 -2.24296251e+00 -1.71227019e+00 -4.82544059e-01
9.42834595e-01 1.98499700e+00 2.21647124e+00 1.54496760e+00
2.43047086e-01 -1.15870834e+00 -2.08583521e+00 -2.16588660e+00
-1.35664344e+00 -7.39685386e+11]
Desvio da solução: 0.009788654225244255

--- Resultados para k = 1.3 ---

Solução aproximada: [-1.57712080e+11 1.72879293e+00 2.75178291e+00 2.65329648e+00
1.47234076e+00 -3.08743727e-01 -1.96407442e+00 -2.81844235e+00
-2.52331818e+00 -1.19924068e+00 6.14014172e-01 2.17662052e+00
2.85208387e+00 2.36367690e+00 9.12075415e-01 -9.11982430e-01
-2.36391317e+00 -2.85181559e+00 -2.17681131e+00 -6.13876722e-01
1.19918091e+00 2.52327859e+00 2.81853485e+00 1.96395875e+00
3.08869812e-01 -1.47241608e+00 -2.65333151e+00 -2.75176469e+00
-1.72840899e+00 1.57712080e+11]
Desvio da solução: 0.0015231075573957826

--- Resultados para k = 1.5 ---

Solução aproximada: [ 1.13863857e+10 2.65508411e+00 4.22754218e+00 4.07571017e+00
2.26203789e+00 -4.74468197e-01 -3.01713242e+00 -4.32961805e+00
-3.87634704e+00 -1.84217436e+00 9.43129005e-01 3.34390006e+00
4.38101092e+00 3.63133708e+00 1.40094228e+00 -1.40089847e+00
-3.63145045e+00 -4.38088950e+00 -3.34396327e+00 -9.43141328e-01
1.84222968e+00 3.87631196e+00 4.32957548e+00 3.01725225e+00
4.74336357e-01 -2.26195029e+00 -4.07575985e+00 -4.22749963e+00
-2.65512692e+00 -1.13863857e+10]
Desvio da solução: 5.613995625619894e-05

--- Resultados para k = 1.8 ---

Solução aproximada: [-1.40758365e+09 4.58802946e+00 7.30514231e+00 7.04287957e+00
3.90873683e+00 -8.19810131e-01 -5.21363424e+00 -7.48163131e+00
-6.69821847e+00 -3.18339250e+00 1.62980180e+00 5.77826019e+00
7.57030468e+00 6.27507225e+00 2.42073440e+00 -2.42073865e+00
-6.27505771e+00 -7.57032948e+00 -5.77823672e+00 -1.62980151e+00
 3.18334906e+00 6.69831668e+00 7.48147303e+00 5.21383480e+00
8.19611869e-01 -3.90858069e+00 -7.04298675e+00 -7.30506028e+00
-4.58811210e+00 1.40758365e+09]
Desvio da solução: 9.283233838012651e-07

--- Resultados para k = 2.0 ---

Solução aproximada: [-1.75889094e+07 6.29364423e+00 1.00206976e+01 9.66112167e+00
5.36163763e+00 -1.12441060e+00 -7.15191025e+00 -1.02627476e+01
-9.18831592e+00 -4.36673667e+00 2.23563282e+00 7.92630585e+00
1.03844956e+01 8.60776987e+00 3.32065092e+00 -3.32065106e+00
-8.60776981e+00 -1.03844947e+01 -7.92630858e+00 -2.23562797e+00
4.36673017e+00 9.18832349e+00 1.02627385e+01 7.15192259e+00
1.12439306e+00 -5.36161370e+00 -9.66115251e+00 -1.00206595e+01
-6.29368891e+00 1.75889094e+07]
Desvio da solução: 7.473162764803892e-08

--- Resultados para k = 2.3 ---

Solução aproximada: [ 8.96295970e+05 9.57188016e+00 1.52402039e+01 1.46933739e+01
8.15437372e+00 -1.71008919e+00 -1.08771537e+01 -1.56083679e+01
-1.39742670e+01 -6.64127358e+00 3.40013079e+00 1.20549078e+01
1.57935320e+01 1.30913304e+01 5.05030584e+00 -5.05030567e+00
-1.30913310e+01 -1.57935307e+01 -1.20549098e+01 -3.40012824e+00
6.64127084e+00 1.39742694e+01 1.56083665e+01 1.08771534e+01
1.71009157e+00 -8.15437840e+00 -1.46933671e+01 -1.52402126e+01
-9.57187025e+00 -8.96295970e+05]
Desvio da solução: 2.981640188754405e-09

--- Resultados para k = 2.6 ---

Solução aproximada: [ 5.96924021e+04 1.38271805e+01 2.20154410e+01 2.12255044e+01
1.17795087e+01 -2.47033115e+00 -1.57127374e+01 -2.25472707e+01
-2.01867142e+01 -9.59373639e+00 4.91170117e+00 1.74140779e+01
2.28147529e+01 1.89112547e+01 7.29548511e+00 -7.29548519e+00
-1.89112545e+01 -2.28147533e+01 -1.74140775e+01 -4.91170155e+00
9.59373680e+00 2.01867137e+01 2.25472714e+01 1.57127364e+01
2.47033252e+00 -1.17795105e+01 -2.12255020e+01 -2.20154439e+01
-1.38271772e+01 -5.96924021e+04]
Desvio da solução: 1.8229304074645285e-10

--- Resultados para k = 3 ---

Solução aproximada: [-2432.07944394 21.24111422 33.8198082 32.60631456
18.0955141 -3.79488809 -24.13768247 -34.63679475
-31.01054241 -14.73775998 7.54528438 26.75125824
35.04769679 29.05119975 11.20721937 -11.20721941
-29.05119965 -35.04769693 -26.75125808 -7.54528451
14.73776007 31.01054238 34.63679471 24.13768258
3.79488788 -18.09551377 -32.60631503 -33.8198076
-21.24111492 2432.07944431]
Desvio da solução: 6.733207399029602e-12

--- Resultados para k = 3.5 ---

Solução aproximada: [ 10.12795706 33.73010329 53.70460236 51.7776202 28.73500604
-6.02614172 -38.32974567 -55.0019474 -49.24359257 -23.40301729
11.98163245 42.48000698 55.65444473 46.13222898 17.79664947
-17.79664946 -46.13222899 -55.65444472 -42.48000699 -11.98163244
23.4030173 49.24359254 55.00194745 38.32974558 6.02614184
-28.7350062 -51.77762001 -53.70460258 -33.73010305 -10.12795718]
Desvio da solução: 1.9268254801781e-13

--- Resultados para k = 4 ---

Solução aproximada: [ 76.06189383 50.34931111 80.16546826 77.28904462 42.89306811
-8.99528823 -57.21524913 -82.10203039 -73.50647241 -34.93394807
17.88511691 63.4103917 83.07602002 68.86210525 26.56525855
-26.56525876 -68.86210463 -83.07602104 -63.41039026 -17.88511878
34.93395043 73.50646952 82.10203382 57.21524519 8.99529261
-42.89307284 -77.28903962 -80.16547344 -50.34930582 -76.0618965 ]
Desvio da solução: 1.5169432972663809e-13

--- Resultados para k = 5 ---

Solução aproximada: [-744.5449948 98.33852192 156.57317982 150.9551416 83.77556899
-17.56898451 -111.74846055 -160.35560551 -143.56725221 -68.23043995
34.93193637 123.8483578 162.2579144 134.49623298 51.88534435
-51.88535429 -134.49620579 -162.25795185 -123.84831919 -34.93196785
68.23045947 143.56724459 160.35560607 111.74845883 17.56899673
-83.77559961 -150.95508833 -156.57325498 -98.33843089 744.54494639]
Desvio da solução: 4.5361408323041287e-13

--- Resultados para k = 6.5 ---

Solução aproximada: [ 306.43778249 216.05340862 343.98712676 331.65307722 184.04980283
-38.59345636 -245.5174064 -352.29487325 -315.42388194 -149.89555163
76.73879331 272.10131567 356.47448265 295.49397416 113.98680268
-113.98728428 -295.49256309 -356.47672979 -272.09836407 -76.74230909
149.89951209 315.4195547 352.29953675 245.51240099 38.59882203
-184.05553397 -331.64700668 -343.9934734 -216.0468818 -306.44107719]
Desvio da solução: 3.049521916353213e-13

--- Resultados para k = 8 ---

Solução aproximada: [-1214.43437347 402.2228963 641.8154133 617.91153044
343.44967843 -72.17289463 -457.59975705 -656.85985093
-588.07518 -279.3932681 142.95920507 507.43813679
664.42802543 551.09657738 212.3051049 -212.2867667
-551.15499695 -664.31983067 -507.61063813 -142.70510944
279.03985542 588.54385595 656.26364009 458.33054389
71.30701755 -342.45581932 -619.01785231 -640.62071354
-403.47424164 1215.06981194]
Desvio da solução: 1.5825023009666053e-12

--- Resultados para k = 10 ---

Solução aproximada: [-3928.00303421 757.3640925 1280.5832034 1181.43105984
694.30294583 -162.32793504 -874.64561916 -1299.72264427
-1134.09012259 -557.94433093 289.32158747 983.03823097
1303.7869318 1072.23256742 416.79915846 -414.66777746
-1078.68679752 -1292.84351036 -998.70578752 -268.67734265
532.10556731 1165.2614733 1263.19561486 916.40832931
115.61772893 -643.12452612 -1236.39148791 -1222.73328406
-817.03104254 3958.14675094]
Desvio da solução: 1.929452497327449e-12

--- Resultados para k = 20 ---

Solução aproximada: [ 10949.53949015 9965.22338565 4375.73770137 16358.10441322
-1390.45833106 4849.11597682 -11824.12490814 -7067.86422699
-11026.00056514 -3559.95172602 2017.67562341 8037.15710563
9927.73542266 9763.23987239 1286.30842939 -451.17826139
-12047.54730064 -6780.4451516 -11295.78554688 676.14320354
1845.50510656 11671.39654165 7324.56871983 10941.05565278
-3579.15786208 -138.00421737 -14602.54888366 -6353.82027105
-7810.54135493 -12061.07803811]
Desvio da solução: 1.521942954329661e-11

--- Resultados para k = 50 ---

Solução aproximada: [ 84632.5364527 62756.36513771 119911.86335453 222972.72443061
75261.81127623 -133980.82572547 75012.1639139 -288096.61070468
-153528.30194208 34653.69265622 -51900.50251902 141050.81829563
169834.34922224 171059.32886203 -28504.54707006 5184.59571784
-125126.05646937 -194457.71439951 -152060.52208958 72942.36211717
-38420.85092719 145027.02808804 287152.06489592 -68627.61472053
158250.18089256 -150197.48182684 -138410.24909811 -135998.62036204
-151014.73343475 -15377.25402412]
Desvio da solução: 2.572689065351729e-10

--- Resultados para k = 100 ---

Solução aproximada: [-1047642.42433127 1198754.37723802 584879.73364253 1618727.09219034
460409.75418747 268745.27850576 -1567576.68607678 -675560.38491809
-1515458.26928068 -243610.86961855 -71699.90648465 1230886.09280524
1245992.81483573 1185378.53731372 26306.47981981 86092.45185094
-1328198.27924379 -1437144.40733872 -603020.15179506 -668215.85645359
661944.69573037 1527405.38040133 495568.31539767 1574010.38353347
107288.69042517 -1235536.41627195 -712076.32468896 -1118337.35989568
-1337567.16472755 1289254.42324774]
Desvio da solução: 4.248925335606488e-09

--- Resultados para k = 200 ---

Solução aproximada: [ -1530667.34465275 2626698.29492198 18065487.27119772
4888817.47732183 6665768.29796466 -3664328.57649286
-6208588.15002503 -4407422.76560693 -15837619.7088785
-6379547.2631297 8744281.72893899 7545645.93484305
7173775.14258692 3980355.90206967 16416481.10826211
-15011409.58571111 -2694177.41155057 -13531190.65030401
-6829804.58961376 -172209.03779668 1858486.28000782
9592732.94746244 10010712.17837579 7997067.67427783
2858205.0394027 -9584690.1866314 -7323408.77124175
-10520038.65885304 -3641040.83419864 -1088371.7429468 ]
Desvio da solução: 3.386067577288218e-08

f) Calcule a solução numérica do problema (1) para os pontos y_1 = (0,0), y_2 = (0.7, 0.4), y_3 = (0, 0.7), y_4 = (-0.7, 0.4), y_5 = (-0.7, -0.4), y_6 = (0, -
0.7) e y_7 = (0.7, -0.4). Basta calcular:ϕ(yi )=∑(j=1)^30 u_ij^ (r)v_j , para i = 1, ..., 7. Os v_j são as componentes do vetor obtido no item (d). Observe que
a conta do u^ agora deve ser feita com os y_i.
In [11]: import numpy as np

# Função para calcular a matriz U para um determinado valor de k

def calculate_U(k):
t1_values = np.linspace(0, 2 * np.pi, 60)
t2_values = np.linspace(0, 2 * np.pi, 30)

P1 = np.array([(np.cos(t), np.sin(t)) for t in t1_values])

 P2 = np.array([k * np.array([np.cos(t), np.sin(t)]) for t in t2_values])

U_matrix = np.zeros((60, 30))

for i in range(60):
for j in range(30):
r = np.linalg.norm(P1[i] - P2[j])
if r < 1e-10:
u_star = 0
else:
u_star = 1 / (2 * np.pi) * np.log(1 / r)
U_matrix[i, j] = u_star

return U_matrix

# Vetor b
t1_values = np.linspace(0, 2 * np.pi, 60)
b = np.sin(3 * t1_values)

# Valores de k fornecidos
k_values = [1.05, 1.1, 1.2, 1.3, 1.5, 1.8, 2.0, 2.3, 2.6, 3, 3.5, 4, 5, 6.5, 8, 10, 20, 50, 100, 200]

# Pontos y_i fornecidos

y_values = [(0, 0), (0.7, 0.4), (0, 0.7), (-0.7, 0.4), (-0.7, -0.4), (0, -0.7), (0.7, -0.4)]

# Conjuntos de valores de ϕ(y_i) para cada valor de k

phi_values_for_k = {}

for k in k_values:
U_matrix_k = calculate_U(k)

# Fatoração QR de U_matrix_k
Q, R = np.linalg.qr(U_matrix_k)

# Resolvendo o sistema triangular superior Rv = QTb usando substituição para frente

Qt_b = np.dot(Q.T, b)
v_approx = np.linalg.solve(R, Qt_b)

# Função para calcular a contribuição de cada ponto y_i

def calculate_phi(y):
contributions = []
for j in range(30):
t2 = np.arctan2(y[1], y[0]) # Cálculo de t2 para o ponto y
t2 = (t2 + 2 * np.pi) % (2 * np.pi) # Ajuste para que t2 esteja no intervalo [0, 2π]
r = np.linalg.norm(y - k * np.array([np.cos(t2), np.sin(t2)]))
if r < 1e-10:
u_star = 0
else:
u_star = 1 / (2 * np.pi) * np.log(1 / r)
contributions.append(u_star * v_approx[j])
return np.sum(contributions)

# Calcular ϕ(y_i) para cada ponto y_i usando as soluções aproximadas correspondentes
phi_values = []
for y in y_values:
phi_y_i = calculate_phi(y)
phi_values.append(phi_y_i)

phi_values_for_k[k] = phi_values

# Exibindo os conjuntos de valores de ϕ(y_i) para cada valor de k

for k, phi_values in phi_values_for_k.items():
print(f"--- Resultados para k = {k} ---")
for i, y in enumerate(y_values):
print(f"ϕ(y_{i+1}) = {phi_values[i]}")
print()
--- Resultados para k = 1.05 ---
ϕ(y_1) = 6.103515625e-05
ϕ(y_2) = -0.00390625
ϕ(y_3) = -0.00390625
ϕ(y_4) = -0.00390625
ϕ(y_5) = -0.00390625
ϕ(y_6) = -0.00390625
ϕ(y_7) = -0.00390625

--- Resultados para k = 1.1 ---

ϕ(y_1) = 0.0
ϕ(y_2) = -6.103515625e-05
ϕ(y_3) = -0.00018310546875
ϕ(y_4) = -6.103515625e-05
ϕ(y_5) = -6.103515625e-05
ϕ(y_6) = -0.00018310546875
ϕ(y_7) = -6.103515625e-05
--- Resultados para k = 1.2 ---
ϕ(y_1) = 3.814697265625e-06
ϕ(y_2) = 0.0
ϕ(y_3) = -3.0517578125e-05
ϕ(y_4) = 0.0
ϕ(y_5) = -1.52587890625e-05
ϕ(y_6) = -3.0517578125e-05
ϕ(y_7) = 0.0

--- Resultados para k = 1.3 ---

ϕ(y_1) = 0.0
ϕ(y_2) = 3.814697265625e-06
ϕ(y_3) = 1.9073486328125e-06
ϕ(y_4) = 3.814697265625e-06
ϕ(y_5) = 3.814697265625e-06
ϕ(y_6) = 1.9073486328125e-06
ϕ(y_7) = 0.0

--- Resultados para k = 1.5 ---

ϕ(y_1) = 5.960464477539062e-07
ϕ(y_2) = -2.384185791015625e-07
ϕ(y_3) = -2.384185791015625e-07
ϕ(y_4) = -1.1920928955078125e-07
ϕ(y_5) = -1.1920928955078125e-07
ϕ(y_6) = -2.384185791015625e-07
ϕ(y_7) = -2.384185791015625e-07

--- Resultados para k = 1.8 ---

ϕ(y_1) = -5.960464477539063e-08
ϕ(y_2) = 0.0
ϕ(y_3) = 0.0
ϕ(y_4) = 0.0
ϕ(y_5) = 0.0
ϕ(y_6) = 0.0
ϕ(y_7) = 0.0

--- Resultados para k = 2.0 ---

ϕ(y_1) = -2.3283064365386963e-10
ϕ(y_2) = -1.1641532182693481e-10
ϕ(y_3) = -2.3283064365386963e-10
ϕ(y_4) = -1.1641532182693481e-10
ϕ(y_5) = -1.1641532182693481e-10
ϕ(y_6) = -2.3283064365386963e-10
ϕ(y_7) = -1.1641532182693481e-10

--- Resultados para k = 2.3 ---

ϕ(y_1) = -1.4551915228366852e-11
ϕ(y_2) = -2.1827872842550278e-11
ϕ(y_3) = -1.4551915228366852e-11
ϕ(y_4) = -2.1827872842550278e-11
ϕ(y_5) = -2.1827872842550278e-11
ϕ(y_6) = -1.4551915228366852e-11
ϕ(y_7) = -2.1827872842550278e-11

--- Resultados para k = 2.6 ---

ϕ(y_1) = -7.275957614183426e-12
ϕ(y_2) = -1.8189894035458565e-12
ϕ(y_3) = -3.637978807091713e-12
ϕ(y_4) = -1.8189894035458565e-12
ϕ(y_5) = -1.8189894035458565e-12
ϕ(y_6) = -3.637978807091713e-12
ϕ(y_7) = -1.8189894035458565e-12

--- Resultados para k = 3 ---

ϕ(y_1) = 1.1368683772161603e-13
ϕ(y_2) = 1.1368683772161603e-13
ϕ(y_3) = 5.684341886080802e-14
ϕ(y_4) = 1.1368683772161603e-13
ϕ(y_5) = 1.1368683772161603e-13
ϕ(y_6) = 5.684341886080802e-14
ϕ(y_7) = 1.1368683772161603e-13

--- Resultados para k = 3.5 ---

ϕ(y_1) = 1.3322676295501878e-15
ϕ(y_2) = 8.43769498715119e-15
ϕ(y_3) = -7.327471962526033e-15
ϕ(y_4) = 8.43769498715119e-15
ϕ(y_5) = 0.0
ϕ(y_6) = -7.327471962526033e-15
ϕ(y_7) = 6.661338147750939e-16
--- Resultados para k = 4 ---
ϕ(y_1) = 3.197442310920451e-14
ϕ(y_2) = 3.552713678800501e-15
ϕ(y_3) = 2.6645352591003757e-14
ϕ(y_4) = 3.552713678800501e-15
ϕ(y_5) = 3.552713678800501e-15
ϕ(y_6) = 2.6645352591003757e-14
ϕ(y_7) = 5.329070518200751e-15

--- Resultados para k = 5 ---

ϕ(y_1) = 0.0
ϕ(y_2) = -2.842170943040401e-14
ϕ(y_3) = 0.0
ϕ(y_4) = -2.842170943040401e-14
ϕ(y_5) = -2.842170943040401e-14
ϕ(y_6) = 0.0
ϕ(y_7) = -2.842170943040401e-14

--- Resultados para k = 6.5 ---

ϕ(y_1) = -2.842170943040401e-14
ϕ(y_2) = 1.4210854715202004e-14
ϕ(y_3) = -9.947598300641403e-14
ϕ(y_4) = 1.4210854715202004e-14
ϕ(y_5) = 1.4210854715202004e-14
ϕ(y_6) = -9.947598300641403e-14
ϕ(y_7) = 1.4210854715202004e-14

--- Resultados para k = 8 ---

ϕ(y_1) = 2.2737367544323206e-13
ϕ(y_2) = 1.1368683772161603e-13
ϕ(y_3) = 5.684341886080802e-14
ϕ(y_4) = 5.684341886080802e-14
ϕ(y_5) = 5.684341886080802e-14
ϕ(y_6) = 5.684341886080802e-14
ϕ(y_7) = 1.1368683772161603e-13

--- Resultados para k = 10 ---

ϕ(y_1) = 0.0
ϕ(y_2) = -2.2737367544323206e-13
ϕ(y_3) = 2.2737367544323206e-13
ϕ(y_4) = -9.094947017729282e-13
ϕ(y_5) = -2.2737367544323206e-13
ϕ(y_6) = 2.2737367544323206e-13
ϕ(y_7) = -9.094947017729282e-13

--- Resultados para k = 20 ---

ϕ(y_1) = 3.637978807091713e-12
ϕ(y_2) = 0.0
ϕ(y_3) = 9.094947017729282e-13
ϕ(y_4) = 0.0
ϕ(y_5) = 0.0
ϕ(y_6) = 9.094947017729282e-13
ϕ(y_7) = 0.0

--- Resultados para k = 50 ---

ϕ(y_1) = 5.093170329928398e-11
ϕ(y_2) = -2.546585164964199e-11
ϕ(y_3) = 1.4551915228366852e-11
ϕ(y_4) = -2.546585164964199e-11
ϕ(y_5) = -2.546585164964199e-11
ϕ(y_6) = 1.4551915228366852e-11
ϕ(y_7) = -2.546585164964199e-11

--- Resultados para k = 100 ---

ϕ(y_1) = -1.1641532182693481e-10
ϕ(y_2) = 9.313225746154785e-10
ϕ(y_3) = 6.984919309616089e-10
ϕ(y_4) = 9.313225746154785e-10
ϕ(y_5) = 9.313225746154785e-10
ϕ(y_6) = 6.984919309616089e-10
ϕ(y_7) = 9.313225746154785e-10

--- Resultados para k = 200 ---

ϕ(y_1) = -6.51925802230835e-09
ϕ(y_2) = 1.280568540096283e-09
ϕ(y_3) = 9.313225746154785e-10
ϕ(y_4) = 1.280568540096283e-09
ϕ(y_5) = 1.280568540096283e-09
ϕ(y_6) = 9.313225746154785e-10
ϕ(y_7) = 1.280568540096283e-09
(g) Compare a resposta nos pontos do item (f) com a solução exata.

(h) Para qual r a solução foi a melhor dentro da sua comparação?

In [12]: import numpy as np

# Função para calcular a matriz U para um determinado valor de k

def calculate_U(k):
t1_values = np.linspace(0, 2 * np.pi, 60)
t2_values = np.linspace(0, 2 * np.pi, 30)

P1 = np.array([(np.cos(t), np.sin(t)) for t in t1_values])

P2 = np.array([k * np.array([np.cos(t), np.sin(t)]) for t in t2_values])

U_matrix = np.zeros((60, 30))

for i in range(60):
for j in range(30):
r = np.linalg.norm(P1[i] - P2[j])
if r < 1e-10:
u_star = 0
else:
u_star = 1 / (2 * np.pi) * np.log(1 / r)
U_matrix[i, j] = u_star

return U_matrix

# Vetor b
t1_values = np.linspace(0, 2 * np.pi, 60)
b = np.sin(3 * t1_values)

# Valores de k fornecidos
k_values = [1.05, 1.1, 1.2, 1.3, 1.5, 1.8, 2.0, 2.3, 2.6, 3, 3.5, 4, 5, 6.5, 8, 10, 20, 50, 100, 200]

# Pontos y_i fornecidos

y_values = [(0, 0), (0.7, 0.4), (0, 0.7), (-0.7, 0.4), (-0.7, -0.4), (0, -0.7), (0.7, -0.4)]

# Conjuntos de valores de ϕ(y_i) para cada valor de k

phi_values_for_k = {}

# Solução exata
def exact_solution(r, theta):
return r**3 * np.sin(3 * theta)

best_r = None
best_phi_values = None
best_errors = None

for k in k_values:
U_matrix_k = calculate_U(k)

# Fatoração QR de U_matrix_k
Q, R = np.linalg.qr(U_matrix_k)

# Resolvendo o sistema triangular superior Rv = QTb usando substituição para frente

Qt_b = np.dot(Q.T, b)
v = np.linalg.solve(R, Qt_b)

# Calculando os valores de phi para cada ponto y_i

def calculate_phi(y):
contributions = []
for j in range(30):
t2 = np.arctan2(y[1], y[0]) # Cálculo de t2 para o ponto y
t2 = (t2 + 2 * np.pi) % (2 * np.pi) # Ajuste para que t2 esteja no intervalo [0, 2π]
r = np.linalg.norm(y - k * np.array([np.cos(t2), np.sin(t2)]))
if r < 1e-10:
u_star = 0
else:
u_star = 1 / (2 * np.pi) * np.log(1 / r)
contributions.append(u_star * v[j])
return np.sum(contributions)

phi_values = []
errors = []
for y in y_values:
phi_y_i = calculate_phi(y)
phi_exact_y_i = exact_solution(np.linalg.norm(y), np.arctan2(y[1], y[0]))
error_i = np.abs(phi_y_i - phi_exact_y_i)
phi_values.append(phi_y_i)
errors.append(error_i)

mean_error = np.mean(errors)
if best_r is None or mean_error < best_error:
 best_error = mean_error
best_r = k
best_phi_values = phi_values
best_errors = errors

# Exibindo os resultados apenas para o melhor valor de k

for i, y in enumerate(y_values):
print(f"Para o ponto y_{i+1}:")
print(f"Valor calculado de ϕ(y_{i+1}) = {best_phi_values[i]}")
print(f"Valor exato de ϕ(y_{i+1}) = {exact_solution(np.linalg.norm(y), np.arctan2(y[1], y[0]))}")
print(f"Erro absoluto = {best_errors[i]}")
print(f"Valor de r = {np.linalg.norm(y)}")
print()

print(f"O valor de k que resulta no menor erro absoluto médio é: {best_r}")

Para o ponto y_1:
Valor calculado de ϕ(y_1) = 3.814697265625e-06
Valor exato de ϕ(y_1) = 0.0
Erro absoluto = 3.814697265625e-06
Valor de r = 0.0

Para o ponto y_2:

Valor calculado de ϕ(y_2) = 0.0
Valor exato de ϕ(y_2) = 0.5239999999999999
Erro absoluto = 0.5239999999999999
Valor de r = 0.8062257748298549

Para o ponto y_3:

Valor calculado de ϕ(y_3) = -3.0517578125e-05
Valor exato de ϕ(y_3) = -0.3429999999999999
Erro absoluto = 0.3429694824218749
Valor de r = 0.7

Para o ponto y_4:

Valor calculado de ϕ(y_4) = 0.0
Valor exato de ϕ(y_4) = 0.5239999999999999
Erro absoluto = 0.5239999999999999
Valor de r = 0.8062257748298549

Para o ponto y_5:

Valor calculado de ϕ(y_5) = -1.52587890625e-05
Valor exato de ϕ(y_5) = -0.5239999999999999
Erro absoluto = 0.5239847412109374
Valor de r = 0.8062257748298549

Para o ponto y_6:

Valor calculado de ϕ(y_6) = -3.0517578125e-05
Valor exato de ϕ(y_6) = 0.3429999999999999
Erro absoluto = 0.3430305175781249
Valor de r = 0.7

Para o ponto y_7:

Valor calculado de ϕ(y_7) = 0.0
Valor exato de ϕ(y_7) = -0.5239999999999999
Erro absoluto = 0.5239999999999999
Valor de r = 0.8062257748298549

O valor de k que resulta no menor erro absoluto médio é: 1.2

In [ ]: