def test_estep_test_solution(self):
post, ll = em.estep(ts.X, ts.mixture_initial)
self.assertEqual(np.allclose(post, ts.post_first_estep),
True, f'Post not as expected:\nGot{post}\nExpected {ts.post_first_estep}')
self.assertAlmostEqual(ll, ts.ll_first_estep, 8,
f'Log likelihood: got {ll}, expected {ts.ll_first_estep}')
def run_estep_test_input(self, test):
X, K, mixture, expected_post, expected_ll = test.data()
post, ll = em.estep(X, mixture)
self.assertEqual(np.allclose(post, expected_post),
True, f'Post not as expected:\nGot {post}\nExpected {expected_post}')
self.assertEqual(np.isclose(ll, expected_ll),
True, f'Log likelihood: got {ll}, expected {expected_ll}')
K = 4
n, d = X.shape
seed = 0
mixture, _ = common.init(X, K, seed)
# print("Input:")
# print('X:\n' + str(X))
# print('K: ' + str(K))
# print('Mu:\n' + str(mixture.mu))
# print('Var: ' + str(mixture.var))
# print('P: ' + str(mixture.p))
# print()
# print("After first E-step:")
post, ll = em.estep(X, mixture)
# print('post:\n' + str(post))
# print('LL:' + str(ll))
# print()
# print("After first M-step:")
mu, var, p = em.mstep(X, post, mixture)
# print('Mu:\n' + str(mu))
# print('Var: ' + str(var))
# print('P: ' + str(p))
# print()
# print("After a run")
(mu, var, p), post, ll = em.run(X, mixture, post)
# print('Mu:\n' + str(mu))
# print('Var: ' + str(var))
X = np.loadtxt("test_incomplete.txt")
X_gold = np.loadtxt("test_complete.txt")
# X = np.loadtxt("netflix_incomplete.txt")
# X_gold = np.loadtxt("netflix_complete.txt")
K = 4
n, d = X.shape
seed = 0
# TODO: Your code here
[mixture, post] = common.init(X, K, seed)
# [m,n] = np.shape(X)
# for i in range(m):
# Cu = np.where(X[i] != 0)[0] # return tuple so need [0]
# Hu = np.where(X[i] == 0)[0]
# print(Cu)
# print(Hu)
print(X)
print(mixture)
print(em.estep(X, mixture))
# print(naive_em.estep(X, mixture))
import numpy as np
import em
import common
X = np.loadtxt("test_incomplete.txt")
X_gold = np.loadtxt("test_complete.txt")
K = 4
n, d = X.shape
seed = 0
initial_mixture, post = common.init(X, K, seed)
print(X)
print(initial_mixture)
post, LL = em.estep(X, initial_mixture)
print(post)
print(LL)
new_mixture = em.mstep(X, post, initial_mixture, 0.25)
print(new_mixture)
import common
# =============================================================================
# 7. Implementing EM for matrix completion
# =============================================================================
X = np.loadtxt("test_incomplete.txt")
X_gold = np.loadtxt("test_complete.txt")
K = 4
n, d = X.shape
seed = 0
# for incomplete case
mixture, post = common.init(X, K=K, seed=seed)
post, log_likelihood = em.estep(X, mixture)
mixture = em.mstep(X, post, mixture)
# =============================================================================
# 8. Using the mixture model for collaborative filtering
# Reporting log likelihood values on Netflix data
# =============================================================================
X = np.loadtxt("netflix_incomplete.txt")
mixture, post = common.init(X, K=1, seed=0)
post, log_likelihood = em.estep(X, mixture)
mixtured = em.mstep(X, post, mixture)
Ks = [1, 12]
seeds = [0, 1, 2, 3, 4]
for K in Ks:
for seed in seeds:
mixture, post = common.init(X, K=K, seed=seed) # Initialize K-means
mixture, post, log_likelihood = naive_em.run(X, mixture, post)
common.plot(X, mixture, post, [K, seed])
print(K, seed, log_likelihood)
# =============================================================================
# 5. Bayesian Information Criterion
# Picking the best K
# =============================================================================
for K in Ks:
mixture, post = common.init(X, K=K) # Initialize K-means
mixture, post, log_likelihood = naive_em.run(X, mixture, post)
BIC = common.bic(X, mixture, log_likelihood)
print(K, BIC)
# =============================================================================
# 7. Implementing EM for matrix completion
# Test for comlete case
# =============================================================================
X = np.loadtxt("toy_data.txt")
Ks = [1, 2, 3, 4]
seeds = [0, 1, 2, 3, 4]
mixture, post = common.init(X, 3, seed=0)
post1, log_likelihood1 = naive_em.estep(X, mixture)
post2, log_likelihood2 = em.estep(X, mixture)
print(log_likelihood1, log_likelihood2)
# print(BIC_best, K_best)
######################### 8. Using the mixture model for collaborative filtering
# X = np.loadtxt("netflix_incomplete.txt")
# X_gold = np.loadtxt('netflix_complete.txt')
X = np.loadtxt("test_incomplete.txt")
X_gold = np.loadtxt('test_complete.txt')
K = 4
seed = 0
[mixture, post] = common.init(X, K, seed)
[post, L] = em.estep(X, mixture)
mixture = em.mstep(X, post, mixture)
print(post)
print(L)
print(mixture)
[mixture, post, L] = em.run(X, mixture, post)
print(post)
print(L)
print(mixture)
X_prep = em.fill_matrix(X, mixture)
print(X_prep)
RMSE = common.rmse(X_gold, X_prep)
print(RMSE)
import naive_em
import common
X = np.loadtxt("test_incomplete.txt")
X_gold = np.loadtxt("test_complete.txt")
K = 4
n, d = X.shape
seed = 0
mixture, post = common.init(X, K, seed)
#post, LL = naive_em.estep(X, mixture)
post, LL = em.estep(X, mixture)
print("After first E-step:")
print("post:")
print(post[0:3,:])
print("LL:", LL)
#mixture = naive_em.mstep(X, post)
mixture = em.mstep(X, post, mixture)
print()
print("After first M-step:")
print("Mu:")
print(mixture.mu)
print("Var:", mixture.var)
print("P:", mixture.p)
def test_estep_toy_dataset(self):
mixture, post = common.init(self.X, 3, 0)
post, ll = em.estep(self.X, mixture)
expected_ll = -1388.0818
self.assertEqual(np.isclose(ll, expected_ll),
True, f'Log likelihood: got {ll}, expected {expected_ll}')