def test_rescale(self):
"""munge -- rescale"""
datamat = [[4, 8, 3, 4, 5, 6, 7, 8, 9], [1, 9, 11, 4, 5, 6, 7, 8, 9],
[3, 2, 3, 99, 5, 6, 7, 8, 9]]
result = munge.rescale(datamat)
expected = [
0.8728715609439696, 0.440225453162812, -0.5773502691896257,
-0.5773502691896257, 5, 6, 7, 8, 9
]
self.assertEqual(expected, result[0])
def test_gradient(self):
"""logistic_regression -- gradient"""
# x is [1, experience, salary]
x = [[1] + row[:2] for row in self.data]
# y is "paid account"
y = [row[2] for row in self.data]
rescaled = munge.rescale(x)
random.seed(0)
x_train, x_test, y_train, y_test = ml.train_test_split(rescaled,
y,
0.33)
beta_0 = [1, 1, 1]
beta_hat = gradient.\
maximize_stochastic(logistic_regression.logistic_log_likelihood_i,
logistic_regression.logistic_log_gradient_i,
x_train,
y_train,
beta_0)
true_positives = false_positives = true_negatives = false_negatives = 0
for x_i, y_i in zip(x_test, y_test):
predict = logistic_regression.logistic(algebra.dot(beta_hat, x_i))
if y_i == 1 and predict >= 0.5:
true_positives += 1
elif y_i == 1:
false_negatives += 1
elif predict >= 0.5:
false_positives += 1
else:
true_negatives += 1
message = "true_pos={0}; false_neg={1}, false_pos={2}; true_neg={3}"
logging.debug(message.format(true_positives,
false_negatives,
false_positives,
true_negatives))
precision = true_positives / float((true_positives + false_positives))
recall = true_positives / float((true_positives + false_negatives))
self.assertEqual(0.93, round(precision, 2))
self.assertEqual(0.82, round(recall, 2))
def test_gradient(self):
"""logistic_regression -- gradient"""
# x is [1, experience, salary]
x = [[1] + row[:2] for row in self.data]
# y is "paid account"
y = [row[2] for row in self.data]
rescaled = munge.rescale(x)
random.seed(0)
x_train, x_test, y_train, y_test = ml.train_test_split(
rescaled, y, 0.33)
beta_0 = [1, 1, 1]
beta_hat = gradient.\
maximize_stochastic(logistic_regression.logistic_log_likelihood_i,
logistic_regression.logistic_log_gradient_i,
x_train,
y_train,
beta_0)
true_positives = false_positives = true_negatives = false_negatives = 0
for x_i, y_i in zip(x_test, y_test):
predict = logistic_regression.logistic(algebra.dot(beta_hat, x_i))
if y_i == 1 and predict >= 0.5:
true_positives += 1
elif y_i == 1:
false_negatives += 1
elif predict >= 0.5:
false_positives += 1
else:
true_negatives += 1
message = "true_pos={0}; false_neg={1}, false_pos={2}; true_neg={3}"
logging.debug(
message.format(true_positives, false_negatives, false_positives,
true_negatives))
precision = true_positives / float((true_positives + false_positives))
recall = true_positives / float((true_positives + false_negatives))
self.assertEqual(0.93, round(precision, 2))
self.assertEqual(0.82, round(recall, 2))
def test_rescale(self):
"""munge -- rescale"""
datamat = [
[4, 8, 3, 4, 5, 6, 7, 8, 9],
[1, 9, 11, 4, 5, 6, 7, 8, 9],
[3, 2, 3, 99, 5, 6, 7, 8, 9]
]
result = munge.rescale(datamat)
expected = [0.8728715609439696,
0.440225453162812,
-0.5773502691896257,
-0.5773502691896257,
5,
6,
7,
8,
9]
self.assertEqual(expected, result[0])