def test_sim():
# This function assesses the performance of the knockoff approach
# relative to its theoretical claims.
np.random.seed(43234)
npos = 30
target_fdr = 0.2
nrep = 10
testers = [[kr.CorrelationEffects(), 300, 100, 6],
[kr.ForwardEffects(pursuit=False), 300, 100, 3.5],
[kr.ForwardEffects(pursuit=True), 300, 100, 3.5],
[kr.OLSEffects(), 3000, 200, 3.5]]
for method in "equi", "sdp":
if method == "sdp" and not has_cvxopt:
continue
for tester_info in testers:
fdr = 0
power = 0
tester = tester_info[0]
n = tester_info[1]
p = tester_info[2]
es = tester_info[3]
for k in range(nrep):
x = np.random.normal(size=(n, p))
x /= np.sqrt(np.sum(x*x, 0))
coeff = es * (-1)**np.arange(npos)
y = np.dot(x[:, 0:npos], coeff) + np.random.normal(size=n)
kn = RegressionFDR(y, x, tester)
tr = kn.threshold(target_fdr)
cp = np.sum(kn.stats >= tr)
cp = max(cp, 1)
fp = np.sum(kn.stats[npos:] >= tr)
fdr += fp/cp
power += np.mean(kn.stats[0:npos] >= tr)
estimated_fdr = (np.sum(kn.stats <= -tr) /
(1 + np.sum(kn.stats >= tr)))
assert_array_equal(estimated_fdr < target_fdr, True)
power /= nrep
fdr /= nrep
assert_array_equal(power > 0.6, True)
assert_array_equal(fdr < target_fdr + 0.05, True)
def test_sim():
# This function assesses the performance of the knockoff approach
# relative to its theoretical claims.
np.random.seed(43234)
npos = 30
target_fdr = 0.2
nrep = 10
testers = [[kr.CorrelationEffects(), 300, 100, 6],
[kr.ForwardEffects(pursuit=False), 300, 100, 3.5],
[kr.ForwardEffects(pursuit=True), 300, 100, 3.5],
[kr.OLSEffects(), 3000, 200, 3.5]]
for method in "equi", "sdp":
if method == "sdp" and not has_cvxopt:
continue
for tester_info in testers:
fdr = 0
power = 0
tester = tester_info[0]
n = tester_info[1]
p = tester_info[2]
es = tester_info[3]
for k in range(nrep):
x = np.random.normal(size=(n, p))
x /= np.sqrt(np.sum(x * x, 0))
coeff = es * (-1)**np.arange(npos)
y = np.dot(x[:, 0:npos], coeff) + np.random.normal(size=n)
kn = RegressionFDR(y, x, tester)
tr = kn.threshold(target_fdr)
cp = np.sum(kn.stats >= tr)
cp = max(cp, 1)
fp = np.sum(kn.stats[npos:] >= tr)
fdr += fp / cp
power += np.mean(kn.stats[0:npos] >= tr)
estimated_fdr = (np.sum(kn.stats <= -tr) /
(1 + np.sum(kn.stats >= tr)))
assert_array_equal(estimated_fdr < target_fdr, True)
power /= nrep
fdr /= nrep
assert_array_equal(power > 0.6, True)
assert_array_equal(fdr < target_fdr + 0.05, True)
def test_testers(p, tester, method):
if method == "sdp" and not has_cvxopt:
return
np.random.seed(2432)
n = 200
y = np.random.normal(size=n)
x = np.random.normal(size=(n, p))
kn = RegressionFDR(y, x, tester, design_method=method)
assert_equal(len(kn.stats), p)
assert_equal(len(kn.fdr), p)
kn.summary() # smoke test
def test_testers(p, tester, method):
if method == "sdp" and not has_cvxopt:
return
np.random.seed(2432)
n = 200
y = np.random.normal(size=n)
x = np.random.normal(size=(n, p))
kn = RegressionFDR(y, x, tester, design_method=method)
assert_equal(len(kn.stats), p)
assert_equal(len(kn.fdr), p)
kn.summary() # smoke test
def test_testers():
# Smoke test
np.random.seed(2432)
n = 200
p = 50
y = np.random.normal(size=n)
x = np.random.normal(size=(n, p))
testers = [
kr.CorrelationEffects(),
kr.ForwardEffects(pursuit=False),
kr.ForwardEffects(pursuit=True),
kr.OLSEffects()
]
for method in "equi", "sdp":
if method == "sdp" and not has_cvxopt:
continue
for tv in testers:
RegressionFDR(y, x, tv, design_method=method)
def test_sim(method, tester, n, p, es):
# This function assesses the performance of the knockoff approach
# relative to its theoretical claims.
if method == "sdp" and not has_cvxopt:
return
np.random.seed(43234)
# Number of variables with a non-zero coefficient
npos = 30
# Aim to control FDR to this level
target_fdr = 0.2
# Number of siumulation replications
nrep = 10
if method == "sdp" and not has_cvxopt:
return
fdr, power = 0, 0
for k in range(nrep):
# Generate the predictors
x = np.random.normal(size=(n, p))
x /= np.sqrt(np.sum(x*x, 0))
# Generate the response variable
coeff = es * (-1)**np.arange(npos)
y = np.dot(x[:, 0:npos], coeff) + np.random.normal(size=n)
kn = RegressionFDR(y, x, tester)
# Threshold to achieve the target FDR
tr = kn.threshold(target_fdr)
# Number of selected coefficients
cp = np.sum(kn.stats >= tr)
# Number of false positives
fp = np.sum(kn.stats[npos:] >= tr)
# Observed FDR
fdr += fp / max(cp, 1)
# Proportion of true positives that are detected
power += np.mean(kn.stats[0:npos] >= tr)
# The estimated FDR may never exceed the target FDR
estimated_fdr = (np.sum(kn.stats <= -tr) /
(1 + np.sum(kn.stats >= tr)))
assert_equal(estimated_fdr < target_fdr, True)
power /= nrep
fdr /= nrep
# Check for reasonable power
assert_array_equal(power > 0.6, True)
# Check that we are close to the target FDR
assert_array_equal(fdr < target_fdr + 0.05, True)
def test_sim(method, tester, n, p, es):
# This function assesses the performance of the knockoff approach
# relative to its theoretical claims.
if method == "sdp" and not has_cvxopt:
return
np.random.seed(43234)
# Number of variables with a non-zero coefficient
npos = 30
# Aim to control FDR to this level
target_fdr = 0.2
# Number of siumulation replications
nrep = 10
if method == "sdp" and not has_cvxopt:
return
fdr, power = 0, 0
for k in range(nrep):
# Generate the predictors
x = np.random.normal(size=(n, p))
x /= np.sqrt(np.sum(x * x, 0))
# Generate the response variable
coeff = es * (-1)**np.arange(npos)
y = np.dot(x[:, 0:npos], coeff) + np.random.normal(size=n)
kn = RegressionFDR(y, x, tester)
# Threshold to achieve the target FDR
tr = kn.threshold(target_fdr)
# Number of selected coefficients
cp = np.sum(kn.stats >= tr)
# Number of false positives
fp = np.sum(kn.stats[npos:] >= tr)
# Observed FDR
fdr += fp / max(cp, 1)
# Proportion of true positives that are detected
power += np.mean(kn.stats[0:npos] >= tr)
# The estimated FDR may never exceed the target FDR
estimated_fdr = (np.sum(kn.stats <= -tr) /
(1 + np.sum(kn.stats >= tr)))
assert_equal(estimated_fdr < target_fdr, True)
power /= nrep
fdr /= nrep
# Check for reasonable power
assert_array_equal(power > 0.6, True)
# Check that we are close to the target FDR
assert_array_equal(fdr < target_fdr + 0.05, True)