def test_unbatched(self):
W1 = np.array([1, -2, 3, 6, 3, -2, 1, 2, 5, 3, 0.5, 1, 1, 1, 1, 1, 1, 1])
T1 = data_dependent_threshhold(W1, fdr = 0.2)
expected = np.abs(W1).min()
self.assertTrue(T1==expected, msg=f'Incorrect data dependent threshhold: T1 should be 0, not {T1}')
W2 = np.array([-1, -2, -3])
T2 = data_dependent_threshhold(W2, fdr = 0.3)
self.assertTrue(T2==np.inf, msg=f'Incorrect data dependent threshhold: T2 should be inf, not {T2}')
W3 = np.array([-5, -1, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10])
T3 = data_dependent_threshhold(W3, fdr = 0.2)
self.assertTrue(T3 == 5, msg=f'Incorrect data dependent threshhold: T3 should be 5, not {T3}')
def Z2selections(Z, groups, q, **kwargs):
# Calculate W statistics
W = kstats.combine_Z_stats(Z, groups, **kwargs)
# Calculate selections
T = kstats.data_dependent_threshhold(W=W, fdr=q)
selected_flags = (W >= T).astype("float32")
return selected_flags, W
def test_batched(self):
W1 = np.array([1]*10)
W2 = np.array([-2, -1, 1, 2, 3, 4, 5, 6, 7, 8])
W3 = np.array([-1]*10)
combined = np.stack([W1, W2, W3]).transpose()
Ts = data_dependent_threshhold(combined, fdr = 0.2)
expected = np.array([1, 2, np.inf])
np.testing.assert_array_almost_equal(
Ts, expected,
err_msg = f"Incorrect data dependent threshhold (batched): Ts should be {expected}, not {Ts}"
)
def check_kstat_fit(
self,
fstat,
fstat_name,
fstat_kwargs={},
min_power=0.8,
max_l2norm=9,
seed=110,
group_features=False,
**sample_kwargs
):
""" fstat should be a class instance inheriting from FeatureStatistic """
# Add defaults to sample kwargs
if 'method' not in sample_kwargs:
sample_kwargs['method'] = 'daibarber2016'
if 'gamma' not in sample_kwargs:
sample_kwargs['gamma'] = 1
if 'n' not in sample_kwargs:
sample_kwargs['n'] = 200
if 'p' not in sample_kwargs:
sample_kwargs['p'] = 50
if 'rho' not in sample_kwargs:
sample_kwargs['rho'] = 0.5
if 'y_dist' not in sample_kwargs:
sample_kwargs['y_dist'] = 'gaussian'
n = sample_kwargs['n']
p = sample_kwargs['p']
rho = sample_kwargs['rho']
y_dist = sample_kwargs['y_dist']
# Create data generating process
np.random.seed(seed)
X, y, beta, _, corr_matrix = graphs.sample_data(**sample_kwargs)
# XtXinv = np.linalg.inv(np.dot(X.T, X))
# Xty = np.dot(X.T, y)
# print(np.dot(XtXinv, Xty))
# Create groups
if group_features:
groups = np.random.randint(1, p+1, size=(p,))
groups = utilities.preprocess_groups(groups)
else:
groups = np.arange(1, p+1, 1)
# Create knockoffs
knockoffs, S = knockadapt.knockoffs.gaussian_knockoffs(
X=X,
groups=groups,
Sigma=corr_matrix,
return_S=True,
verbose=False,
sdp_verbose=False,
S = (1-rho)*np.eye(p)
)
knockoffs = knockoffs[:, :, 0]
# Fit and extract coeffs/T
fstat.fit(
X,
knockoffs,
y,
groups=groups,
**fstat_kwargs,
)
W = fstat.W
T = data_dependent_threshhold(W, fdr = 0.2)
# Test L2 norm
m = np.unique(groups).shape[0]
if m == p:
pair_W = W
else:
pair_W = kstats.combine_Z_stats(fstat.Z, pair_agg='cd')
l2norm = np.power(pair_W - np.abs(beta), 2)
l2norm = l2norm.mean()
self.assertTrue(l2norm < max_l2norm,
msg = f'{fstat_name} fits {y_dist} data very poorly (l2norm = {l2norm} btwn real {beta} / fitted {pair_W} coeffs)'
)
# Test power for non-grouped setting.
# (For group setting, power will be much lower.)
selections = (W >= T).astype('float32')
group_nnulls = utilities.fetch_group_nonnulls(beta, groups)
power = ((group_nnulls != 0)*selections).sum()/np.sum(group_nnulls != 0)
fdp = ((group_nnulls == 0)*selections).sum()/max(np.sum(selections), 1)
self.assertTrue(
power >= min_power,
msg = f"Power {power} for {fstat_name} in equicor case (n={n},p={p},rho={rho}, y_dist {y_dist}, grouped={group_features}) should be > {min_power}. W stats are {W}, beta is {beta}"
)