def test_bbsr_three_good_predictors(self):
x = PREDICTORS_Z.copy()
y = scipy.stats.zscore(np.sum(PREDICTORS, axis=1).flatten())
pp = np.array([1, 1, 2])
weights = np.array([10, 10, 10])
max_k = 3
result = bayes_stats.bbsr(x, y, pp, weights, max_k)
pp = np.array([1, 1, 1])
betas = np.array([0.583039, 0.491769, 0.478452])
betas_resc = np.array([1.0, 1.0, 1.0])
np.testing.assert_array_almost_equal(result['pp'], pp)
np.testing.assert_array_almost_equal(result['betas'], betas)
np.testing.assert_array_almost_equal(result['betas_resc'], betas_resc)
result_2 = bayes_stats.bbsr(x,
y,
pp,
weights,
max_k,
ordinary_least_squares=True)
np.testing.assert_array_almost_equal(result['pp'], result_2['pp'])
np.testing.assert_array_almost_equal(result['betas'],
result_2['betas'])
np.testing.assert_array_almost_equal(result['betas_resc'],
result_2['betas_resc'])
def regression_maker(j, x, y, pp, weights):
level = 0 if j % 100 == 0 else 2
utils.Debug.allprint(base_regression.PROGRESS_STR.format(gn=genes[j],
i=j,
total=G),
level=level)
data = bayes_stats.bbsr(x, utils.scale_vector(y), pp[j, :].flatten(),
weights[j, :].flatten(), nS)
data['ind'] = j
return j, data
def regression_maker(j):
level = 0 if j % 100 == 0 else 2
utils.Debug.allprint(base_regression.PROGRESS_STR.format(gn=self.genes[j], i=j, total=self.G),
level=level)
data = bayes_stats.bbsr(self.X.values,
self.Y.iloc[j, :].values.flatten(),
self.pp.iloc[j, :].values.flatten(),
self.weights_mat.iloc[j, :].values.flatten(),
self.nS)
data['ind'] = j
return data
def test_bbsr(self):
# test when pp.sum() != 0
X = np.array([[1, 0, 1], [2, 1, 0], [1, 1, 1], [0, 0, 1], [2, 1, 2]])
y = np.array([0, 1, 0])
pp = np.array([0, 1, 2, 1, 0])
weights = np.array([1, 0, 2, 1, 5])
max_k = 10
result = bayes_stats.bbsr(X, y, pp, weights, max_k)
pp = np.array([0, 1, 1, 1, 0])
betas = np.array([0.0, 0.0, 0.0])
betas_resc = np.array([0.0, 0.0, 0.0])
dict = {'pp': pp, 'betas': betas, 'betas_resc': betas_resc}
np.testing.assert_equal(result, dict)
def test_bbsr_3(self):
# test when betas and betas_resc are not zero
X = np.array([[1, 3, 1], [2, 1, 0], [1, 10, 5], [2, 6, 1], [2, 1, 8]])
y = np.array([2, 1, 4])
pp = np.array([10, 3, 1, 5, 4])
weights = np.array([10, 10, 10, 10, 10])
max_k = 3
result = bayes_stats.bbsr(X, y, pp, weights, max_k)
pp = np.array([1, 0, 0, 1, 1])
betas = ([0.0, 0.0, 0.53623188])
betas_resc = ([0.0, 0.0, 0.83820926])
check = {'pp': pp, 'betas': betas, 'betas_resc': betas_resc}
for component in check.keys():
for idx in range(0, len(check[component])):
np.testing.assert_array_almost_equal(result[component][idx],
check[component][idx], 2)
def regression_maker(j):
level = 0 if j % 100 == 0 else 2
utils.Debug.allprint(base_regression.PROGRESS_STR.format(
gn=self.genes[j], i=j, total=self.G),
level=level)
data = bayes_stats.bbsr(
self.X.values,
utils.scale_vector(
self.Y.get_gene_data(j, force_dense=True, flatten=True)),
self.pp.iloc[j, :].values.flatten(),
self.weights_mat.iloc[j, :].values.flatten(),
self.nS,
ordinary_least_squares=self.ols_only)
data['ind'] = j
return data
