def test_coxph():
"""
Check that Cox results agree with R
"""
tol = 1.e-5
R_code = """
library(selectiveInference)
set.seed(43)
n = 50
p = 10
sigma = 1.1
x = matrix(rnorm(n*p),n,p)
x=scale(x,TRUE,TRUE)
beta = c(3,2,rep(0,p-2))
tim = as.vector(x%*%beta + sigma*rnorm(n))
tim= tim-min(tim)+1
status=sample(c(0,1),size=n,replace=T)
# first run glmnet
gfit = glmnet(x,Surv(tim,status),standardize=FALSE,family="cox", thresh=1.e-14)
# extract coef for a given lambda; note the 1/n factor!
lambda = 1.5
beta_hat = as.numeric(coef(gfit, s=lambda/n, exact=TRUE))
# compute fixed lambda p-values and selection intervals
out = fixedLassoInf(x,tim,beta_hat,lambda,status=status,family="cox")
pval = out$pv
vars_cox = out$var
"""
rpy.r(R_code)
R_pvals = np.asarray(rpy.r('pval'))
selected_vars = np.asarray(rpy.r('vars_cox'))
tim = np.asarray(rpy.r('tim'))
tim = tim.reshape(-1)
status = np.asarray(rpy.r('status'))
status = status.reshape(-1)
beta_hat = np.asarray(rpy.r('beta_hat'))
x = np.asarray(rpy.r('x'))
L = lasso.coxph(x, tim, status, 1.5)
beta2 = L.fit()
G1 = L.loglike.gradient(beta_hat)
G2 = L.loglike.gradient(beta2)
print(G1, 'glmnet')
print(G2, 'regreg')
yield np.testing.assert_equal, np.array(L.active) + 1, selected_vars
yield np.testing.assert_allclose, beta2, beta_hat, tol, tol, False, 'cox coeff'
yield np.testing.assert_allclose, L.summary(
'onesided')['pval'], R_pvals, tol, tol, False, 'cox pvalues'
def test_coxph():
tol = 1.e-5
R_code = """
library(selectiveInference)
set.seed(43)
n = 50
p = 10
sigma = 1.1
x = matrix(rnorm(n*p),n,p)
x=scale(x,TRUE,TRUE)
beta = c(3,2,rep(0,p-2))
tim = as.vector(x%*%beta + sigma*rnorm(n))
tim= tim-min(tim)+1
status=sample(c(0,1),size=n,replace=T)
# first run glmnet
gfit = glmnet(x,Surv(tim,status),standardize=FALSE,family="cox", thresh=1.e-14)
# extract coef for a given lambda; note the 1/n factor!
lambda = 1.5
beta_hat = as.numeric(coef(gfit, s=lambda/n, exact=TRUE))
# compute fixed lambda p-values and selection intervals
out = fixedLassoInf(x,tim,beta_hat,lambda,status=status,family="cox")
pval = out$pv
vars_cox = out$var
"""
rpy.r(R_code)
R_pvals = np.asarray(rpy.r('pval'))
selected_vars = np.asarray(rpy.r('vars_cox'))
tim = np.asarray(rpy.r('tim'))
tim = tim.reshape(-1)
status = np.asarray(rpy.r('status'))
status = status.reshape(-1)
beta_hat = np.asarray(rpy.r('beta_hat'))
x = np.asarray(rpy.r('x'))
L = lasso.coxph(x, tim, status, 1.5)
beta2 = L.fit()
G1 = L.loglike.gradient(beta_hat)
G2 = L.loglike.gradient(beta2)
print(G1, 'glmnet')
print(G2, 'regreg')
yield np.testing.assert_equal, L.active + 1, selected_vars
yield np.testing.assert_allclose, beta2, beta_hat, tol, tol, False, 'cox coeff'
yield np.testing.assert_allclose, L.summary('onesided')['pval'], R_pvals, tol, tol, False, 'cox pvalues'
def test_coxph():
Q = rr.identity_quadratic(0.01, 0, np.ones(5), 0)
X = np.random.standard_normal((100, 5))
T = np.random.standard_exponential(100)
S = np.random.binomial(1, 0.5, size=(100, ))
L = lasso.coxph(X, T, S, 0.1, quadratic=Q)
L.fit()
L = lasso.coxph(X, T, S, 0.1, quadratic=Q)
L.fit()
C = L.constraints
np.testing.assert_array_less( \
np.dot(L.constraints.linear_part, L.onestep_estimator),
L.constraints.offset)
P = L.summary()['pval']
return L, C, P
def test_coxph():
Q = rr.identity_quadratic(0.01, 0, np.ones(5), 0)
X = np.random.standard_normal((100,5))
T = np.random.standard_exponential(100)
S = np.random.binomial(1, 0.5, size=(100,))
L = lasso.coxph(X, T, S, 0.1, quadratic=Q)
L.fit()
L = lasso.coxph(X, T, S, 0.1, quadratic=Q)
L.fit()
C = L.constraints
np.testing.assert_array_less( \
np.dot(L.constraints.linear_part, L.onestep_estimator),
L.constraints.offset)
P = L.summary()['pval']
return L, C, P
def test_coxph():
Q = identity_quadratic(0.01, 0, np.ones(5), 0)
X = np.random.standard_normal((100,5))
T = np.random.standard_exponential(100)
S = np.random.binomial(1, 0.5, size=(100,))
L = lasso.coxph(X, T, S, 0.1, quadratic=Q)
L.fit()
C = L.constraints
np.testing.assert_array_less( \
np.dot(L.constraints.linear_part, L._onestep),
L.constraints.offset)
I = L.intervals
P = L.active_pvalues
return L, C, I, P