def execute_dual_fs(test_snps, pheno, G0, covar):
"""
implementation of dual-kernel feature selection
"""
result = {}
fs_result = {}
# extract data
G_test = test_snps.read().standardize().val
G_train_unnorm = G0.read().standardize().val
# fs conditioned on full kernel
select = FeatureSelectionInSample(max_log_k=7, order_by_lmm=True)
fs_result["insample_cond_full"] = select.run_select(G_train_unnorm, G_train_unnorm, pheno["vals"], cov=covar["vals"])
best_k, fs_idx, best_mix, best_delta = fs_result["insample_cond_full"]
print "best_k:", best_k, ", best_mix:", best_mix
# set up foreground kernel
G1 = G0[:,fs_idx]
result["full_fs_low"] = single_snp(test_snps, pheno, G0=G0, covar=covar, G1=G1, mixing=best_mix).sort(["Chr", "ChrPos"])["PValue"].as_matrix()
return result, fs_result
def execute_dual_fs(test_snps, pheno, G0, covar):
"""
implementation of dual-kernel feature selection
"""
result = {}
fs_result = {}
# extract data
G_test = test_snps.read().standardize().val
G_train_unnorm = G0.read().standardize().val
# fs conditioned on full kernel
select = FeatureSelectionInSample(max_log_k=7, order_by_lmm=True)
fs_result["insample_cond_full"] = select.run_select(G_train_unnorm,
G_train_unnorm,
pheno["vals"],
cov=covar["vals"])
best_k, fs_idx, best_mix, best_delta = fs_result["insample_cond_full"]
print "best_k:", best_k, ", best_mix:", best_mix
# set up foreground kernel
G1 = G0[:, fs_idx]
result["full_fs_low"] = single_snp(test_snps,
pheno,
G0=G0,
covar=covar,
G1=G1,
mixing=best_mix).sort(
["Chr",
"ChrPos"])["PValue"].as_matrix()
return result, fs_result
def test_regression_lr(self):
# invoke fs
select = FeatureSelectionInSample(n_folds=2, max_log_k=6, order_by_lmm=False, measure="mse", random_state=42)
best_k, feat_idx, best_mix, best_delta = select.run_select(self.G, self.G, self.y, cov=self.G_cov)
# print results
print "best_k:", best_k
print "best_mix:", best_mix
print "best_delta:", best_delta
self.assertEqual(best_k, 32)
self.assertAlmostEqual(best_mix, 0.6786566031577429, places=6)
self.assertAlmostEqual(best_delta, 0.70148446599200931, places=6)
def test_regression_lr(self):
# invoke fs
select = FeatureSelectionInSample(n_folds=2, max_log_k=6, order_by_lmm=False, measure="mse", random_state=42)
best_k, feat_idx, best_mix, best_delta = select.run_select(self.G, self.G, self.y, cov=self.G_cov)
# print results
print("best_k:", best_k)
print("best_mix:", best_mix)
print("best_delta:", best_delta)
self.assertEqual(best_k, 32)
self.assertAlmostEqual(best_mix, 0.6786566031577429, places=6)
self.assertAlmostEqual(best_delta, 0.70148446599200931, places=6)
def test_regression_lmm(self):
# invoke fs
select = FeatureSelectionInSample(n_folds=2, max_log_k=6, order_by_lmm=True, measure="mse", random_state=42)
best_k, feat_idx, best_mix, best_delta = select.run_select(self.G, self.G, self.y, cov=self.G_cov)
# print results
print "best_k:", best_k
print "best_mix:", best_mix
print "best_delta:", best_delta
self.assertEqual(best_k, 64)
self.assertAlmostEqual(best_mix, 0.8621642030968627, places=6)
self.assertAlmostEqual(best_delta, 0.7255878551207211, places=6)
def test_regression_lmm(self):
# invoke fs
select = FeatureSelectionInSample(n_folds=2, max_log_k=6, order_by_lmm=True, measure="mse", random_state=42)
best_k, feat_idx, best_mix, best_delta = select.run_select(self.G, self.G, self.y, cov=self.G_cov)
# print results
print("best_k:", best_k)
print("best_mix:", best_mix)
print("best_delta:", best_delta)
self.assertEqual(best_k, 64)
self.assertAlmostEqual(best_mix, 0.8621642030968627, places=6)
self.assertAlmostEqual(best_delta, 0.7255878551207211, places=6)
def test_old(self):
do_plot = False
from fastlmm.feature_selection.feature_selection_two_kernel import FeatureSelectionInSample
from pysnptools.util import intersect_apply
logging.info("TestSingleSnpAllPlusSelect test_old")
bed_fn = self.pythonpath + "/tests/datasets/synth/all.bed"
pheno_fn = self.pythonpath + "/tests/datasets/synth/pheno_10_causals.txt"
cov_fn = self.pythonpath + "/tests/datasets/synth/cov.txt"
#load data
###################################################################
snp_reader = Bed(bed_fn, count_A1=False)
pheno = Pheno(pheno_fn)
cov = Pheno(cov_fn)
# intersect sample ids
snp_reader, pheno, cov = intersect_apply([snp_reader, pheno, cov])
# read in snps
# partition snps on chr5 vs rest
test_chr = 5
G0 = snp_reader[:, snp_reader.pos[:, 0] != test_chr].read(
order='C').standardize()
test_snps = snp_reader[:, snp_reader.pos[:, 0] == test_chr].read(
order='C').standardize()
y = pheno.read().val[:, 0]
y -= y.mean()
y /= y.std()
# load covariates
X_cov = cov.read().val
X_cov.flags.writeable = False
# invoke feature selection to learn which SNPs to use to build G1
logging.info(
"running feature selection conditioned on background kernel")
# partition data into the first 50 SNPs on chr1 and all but chr1
select = FeatureSelectionInSample(max_log_k=7,
n_folds=7,
order_by_lmm=True,
measure="ll",
random_state=42)
best_k, feat_idx, best_mix, best_delta = select.run_select(G0.val,
G0.val,
y,
cov=X_cov)
# plot out of sample error
if do_plot: select.plot_results(measure="ll")
# select.plot_results(measure="mse")
# print results
logging.info("best_k:{0}".format(best_k))
logging.info("best_mix:{0}".format(best_mix))
logging.info("best_delta:{0}".format(best_delta))
###############################
# use selected SNPs to build G1
logging.info(feat_idx)
G1 = G0[:, feat_idx]
output_file_name = self.file_name("old")
results_df = single_snp(test_snps,
pheno,
G0=G0,
G1=G1,
mixing=best_mix,
h2=None,
leave_out_one_chrom=False,
output_file_name=output_file_name,
count_A1=False)
logging.info("results:")
logging.info("#" * 40)
logging.info(results_df.head())
self.compare_files(results_df, "old")
def test_old(self):
do_plot = False
from fastlmm.feature_selection.feature_selection_two_kernel import FeatureSelectionInSample
from pysnptools.util import intersect_apply
logging.info("TestSingleSnpAllPlusSelect test_old")
bed_fn = self.pythonpath + "/tests/datasets/synth/all.bed"
pheno_fn = self.pythonpath + "/tests/datasets/synth/pheno_10_causals.txt"
cov_fn = self.pythonpath + "/tests/datasets/synth/cov.txt"
#load data
###################################################################
snp_reader = Bed(bed_fn)
pheno = Pheno(pheno_fn)
cov = Pheno(cov_fn)
# intersect sample ids
snp_reader, pheno, cov = intersect_apply([snp_reader, pheno, cov])
# read in snps
# partition snps on chr5 vs rest
test_chr = 5
G0 = snp_reader[:,snp_reader.pos[:,0] != test_chr].read(order='C').standardize()
test_snps = snp_reader[:,snp_reader.pos[:,0] == test_chr].read(order='C').standardize()
y = pheno.read().val[:,0]
y -= y.mean()
y /= y.std()
# load covariates
X_cov = cov.read().val
X_cov.flags.writeable = False
# invoke feature selection to learn which SNPs to use to build G1
logging.info("running feature selection conditioned on background kernel")
# partition data into the first 50 SNPs on chr1 and all but chr1
select = FeatureSelectionInSample(max_log_k=7, n_folds=7, order_by_lmm=True, measure="ll", random_state=42)
best_k, feat_idx, best_mix, best_delta = select.run_select(G0.val, G0.val, y, cov=X_cov)
# plot out of sample error
if do_plot: select.plot_results(measure="ll")
# select.plot_results(measure="mse")
# print results
logging.info("best_k:{0}".format(best_k))
logging.info("best_mix:{0}".format(best_mix))
logging.info("best_delta:{0}".format(best_delta))
###############################
# use selected SNPs to build G1
logging.info(feat_idx)
G1 = G0[:,feat_idx]
output_file_name = self.file_name("old")
results_df = single_snp(test_snps, pheno, G0=G0, G1=G1, mixing=best_mix, h2=None,leave_out_one_chrom=False,output_file_name=output_file_name)
logging.info("results:")
logging.info("#"*40)
logging.info(results_df.head())
self.compare_files(results_df,"old")
