def _fasttwoK(self,force_low_rank,GB_goal):
train_idx = np.r_[10:self.snpreader_whole.iid_count] # iids 10 and on
test_idx = np.r_[0:10] # the first 10 iids
G0_train = self.snpreader_whole[train_idx,:]
G1_train = SnpData(iid=G0_train.iid,sid=[item+"_1" for item in G0_train.sid],val=G0_train.read().val,pos=G0_train.pos,name="Different SNP names for {0}".format(G0_train))
covariate_train = self.covariate_whole[train_idx,:]
pheno_train = self.pheno_whole[train_idx,:]
logging.info("force_low_rank = {0}".format(force_low_rank))
fastlmm1 = FastLMM(force_low_rank=force_low_rank,GB_goal=GB_goal).fit(K0_train=G0_train, K1_train=G1_train, X=covariate_train, y=pheno_train, mixing=.1)
filename = self.tempout_dir + "/model_fasttwoK.flm.p"
pstutil.create_directory_if_necessary(filename)
joblib.dump(fastlmm1, filename)
fastlmm2 = joblib.load(filename)
# predict on test set
G0_test = self.snpreader_whole[test_idx,:]
G1_test = SnpData(iid=G0_test.iid,sid=[item+"_1" for item in G0_test.sid],val=G0_test.read().val,pos=G0_test.pos,name="Different SNP names for {0}".format(G0_test))
covariate_test = self.covariate_whole[test_idx,:]
predicted_pheno, covar = fastlmm2.predict(K0_whole_test=G0_test, K1_whole_test=G1_test, X=covariate_test,count_A1=False)
output_file = self.file_name("fasttwoK"+("_force_low" if force_low_rank else "")+("GB{0}".format(GB_goal) if GB_goal is not None else ""))
Dat.write(output_file,predicted_pheno)
pheno_actual = self.pheno_whole[test_idx,:].read().val[:,0]
#pylab.plot(pheno_actual, predicted_pheno.val,".")
#pylab.show()
self.compare_files(predicted_pheno,"one")
def test_kernel_one(self):
logging.info("TestLmmTrain test_kernel_one")
train_idx = np.r_[10:self.snpreader_whole.iid_count] # iids 10 and on
test_idx = np.r_[0:10] # the first 10 iids
K0_train = SnpKernel(self.snpreader_whole[train_idx,:],standardizer=Unit())
covariate_train = self.covariate_whole[train_idx,:]
pheno_train = self.pheno_whole[train_idx,:]
assert np.array_equal(K0_train.iid,covariate_train.iid), "Expect iids to be the same (so that early and late Unit standardization will give the same result)"
assert np.array_equal(K0_train.iid,pheno_train.iid), "Expect iids to be the same (so that early and late Unit standardization will give the same result)"
fastlmm1 = FastLMM(GB_goal=2).fit(K0_train=K0_train, X=covariate_train, y=pheno_train)
filename = self.tempout_dir + "/model_kernel_one.flm.p"
pstutil.create_directory_if_necessary(filename)
joblib.dump(fastlmm1, filename)
fastlmm2 = joblib.load(filename)
# predict on test set
G0_test = self.snpreader_whole[test_idx,:]
covariate_test = self.covariate_whole[test_idx,:]
predicted_pheno, covar = fastlmm2.predict(K0_whole_test=G0_test, X=covariate_test,count_A1=False)
output_file = self.file_name("kernel_one")
Dat.write(output_file,predicted_pheno)
pheno_actual = self.pheno_whole[test_idx,:].read().val[:,0]
#pylab.plot(pheno_actual, predicted_pheno.val,".")
#pylab.show()
self.compare_files(predicted_pheno,"one") #Expect same results as SNPs "one"
def test_twoK(self):
logging.info("TestLmmTrain test_twoK")
train_idx = np.r_[10:self.snpreader_whole.iid_count] # iids 10 and on
test_idx = np.r_[0:10] # the first 10 iids
G0_train = self.snpreader_whole[train_idx,:]
covariate_train = self.covariate_whole[train_idx,:]
pheno_train = self.pheno_whole[train_idx,:]
fastlmm1 = FastLMM(GB_goal=2).fit(K0_train=G0_train, K1_train=G0_train, X=covariate_train, y=pheno_train)
filename = self.tempout_dir + "/model_one.flm.p"
pstutil.create_directory_if_necessary(filename)
joblib.dump(fastlmm1, filename)
fastlmm2 = joblib.load(filename)
# predict on test set
G0_test = self.snpreader_whole[test_idx,:]
covariate_test = self.covariate_whole[test_idx,:]
predicted_pheno, covar = fastlmm2.predict(K0_whole_test=G0_test, K1_whole_test=G0_test, X=covariate_test,count_A1=False)
output_file = self.file_name("one")
Dat.write(output_file,predicted_pheno)
pheno_actual = self.pheno_whole[test_idx,:].read().val[:,0]
#pylab.plot(pheno_actual, predicted_pheno.val,".")
#pylab.show()
self.compare_files(predicted_pheno,"one")
def test_lr_no_K0(self):
logging.info("TestLinRegTrain test_lr_no_k0")
train_idx = np.r_[10:self.snpreader_whole.iid_count] # iids 10 and on
test_idx = np.r_[0:10] # the first 10 iids
covariate_train3 = self.covariate_whole[train_idx,:].read()
covariate_train3.val = np.array([[float(num)] for num in xrange(covariate_train3.iid_count)])
pheno_train3 = self.pheno_whole[train_idx,:].read()
np.random.seed(0)
pheno_train3.val = covariate_train3.val * 2.0 + 100 + np.random.normal(size=covariate_train3.val.shape) # y = 2*x+100+normal(0,1)
#Learn model, save, load
fastlmm3x = FastLMM(GB_goal=2).fit(X=covariate_train3, y=pheno_train3)
filename = self.tempout_dir + "/model3.flm.p"
joblib.dump(fastlmm3x, filename)
fastlmm3 = joblib.load(filename)
#Predict with model (test on train)
predicted_pheno, covariance = fastlmm3.predict(K0_whole_test=KernelIdentity(pheno_train3.iid), X=covariate_train3,count_A1=False) #test on train
output_file = self.file_name("lr_no_k0")
Dat.write(output_file,predicted_pheno)
self.compare_files(predicted_pheno,"lr_no_k0")
def test_str(self):
logging.info("TestLmmTrain test_str")
G0_train = self.pythonpath + "/tests/datasets/synth/all"
covariate_train = None
pheno_train = self.pythonpath + "/tests/datasets/synth/pheno_10_causals.txt"
fastlmm1 = FastLMM(GB_goal=2).fit(K0_train=G0_train, X=covariate_train, y=pheno_train,count_A1=False)
filename = self.tempout_dir + "/model_str.flm.p"
pstutil.create_directory_if_necessary(filename)
joblib.dump(fastlmm1, filename)
fastlmm2 = joblib.load(filename)
# predict on same
G0_test = G0_train
covariate_test = covariate_train
predicted_pheno, covar = fastlmm2.predict(K0_whole_test=G0_test, X=covariate_test,count_A1=False)
output_file = self.file_name("str")
Dat.write(output_file,predicted_pheno)
#pheno_actual = self.pheno_whole[test_idx,:].read().val[:,0]
#pylab.plot(pheno_actual, predicted_pheno.val,".")
#pylab.show()
self.compare_files(predicted_pheno,"str")
def test_c_reader_dat(self):
snpreader = Dat(self.currentFolder + "/examples/toydata.dat")[:, ::100]
_fortesting_JustCheckExists().input(snpreader)
snpdata1 = snpreader.read()
self.assertEqual(np.float64, snpdata1.val.dtype)
self.assertTrue(
np.allclose(self.snps[:, ::100],
snpdata1.val,
rtol=1e-05,
atol=1e-05))
snpdata1.val[
1,
2] = np.NaN # Inject a missing value to test writing and reading missing values
output = "tempdir/snpreader/toydata.dat"
create_directory_if_necessary(output)
Dat.write(output, snpdata1)
snpdata2 = Dat(output).read()
np.testing.assert_array_almost_equal(snpdata1.val,
snpdata2.val,
decimal=10)
snpdata3 = snpdata1[:, 0:0].read() #create snpdata with no sids
output = "tempdir/snpreader/toydata3.dat"
Dat.write(output, snpdata3)
snpdata4 = Dat(output).read()
assert snpdata3 == snpdata4
def test_api(self):
train_idx = np.r_[10:self.snpreader_whole.iid_count] # iids 10 and on
test_idx = np.r_[0:10] # the first 10 iids
#####################################################
# Train and standardize cov and then apply to test
#####################################################
cov_train, unit_trained = self.covariate_whole[train_idx,:].read().standardize(Unit(),return_trained=True)
cov_test = self.covariate_whole[test_idx,:].read().standardize(unit_trained)
#####################################################
# standardize whole kernel from snps (both ways) and then pull out the 3 parts
#####################################################
whole_kernel = SnpKernel(self.covariate_whole,Unit()).read().standardize(DiagKtoN())
train_kernel = whole_kernel[train_idx].read(order='A',view_ok=True)
test_kernel = whole_kernel[train_idx,test_idx].read(order='A',view_ok=True)
test_test_kernel = whole_kernel[test_idx,test_idx].read(order='A',view_ok=True)
#####################################################
# create train_train, train_test, and test_test based on just the training snps (both standardizations)
#####################################################
K_train = SnpKernel(self.snpreader_whole[train_idx,:],Unit(),block_size=100)
train_train_kernel, snp_trained, kernel_trained = K_train._read_with_standardizing(to_kerneldata=True, kernel_standardizer=DiagKtoN(), return_trained=True)
K_whole_test = _SnpWholeTest(train=self.snpreader_whole[train_idx,:],test=self.snpreader_whole[test_idx,:],standardizer=snp_trained,block_size=100)
train_idx2 = K_whole_test.iid0_to_index(self.snpreader_whole.iid[train_idx]) #The new reader may have the iids in a different order than the original reader
train_test_kernel = K_whole_test[train_idx2,:].read().standardize(kernel_trained)
test_idx2 = K_whole_test.iid0_to_index(self.snpreader_whole.iid[test_idx])
test_test_kernel = K_whole_test[test_idx2,:].read().standardize(kernel_trained)
#####################################################
# How does predict look with whole_test as input?
#####################################################
# a. - standardize whole up front
whole_kernel = SnpKernel(self.snpreader_whole,Unit(),block_size=100).read().standardize()
train_kernel = whole_kernel[train_idx].read(order='A',view_ok=True)
whole_test_kernel = whole_kernel[:,test_idx].read(order='A',view_ok=True)
fastlmm1 = FastLMM(snp_standardizer=SS_Identity(), kernel_standardizer=KS_Identity())
fastlmm1.fit(K0_train=train_kernel, X=self.covariate_whole, y=self.pheno_whole) #iid intersection means we won't really be using whole covar or pheno
predicted_pheno, covar = fastlmm1.predict(K0_whole_test=whole_test_kernel, X=self.covariate_whole,count_A1=False)
output_file = self.file_name("whole")
Dat.write(output_file,predicted_pheno)
self.compare_files(predicted_pheno,"whole")
# b -- just files
fastlmm2 = FastLMM()
fastlmm2.fit(K0_train=self.snpreader_whole[train_idx,:], X=self.covariate_whole, y=self.pheno_whole[train_idx,:]) #iid intersection means we won't really be using whole covar
predicted_pheno, covar = fastlmm2.predict(K0_whole_test=self.snpreader_whole[test_idx,:], X=self.covariate_whole,count_A1=False)
self.compare_files(predicted_pheno,"one")
def test_lr(self):
import matplotlib.pyplot as plt
import pylab
logging.info("TestLmmTrain test_lr")
train_idx = np.r_[10:self.snpreader_whole.iid_count] # iids 10 and on
test_idx = np.r_[0:10] # the first 10 iids
G0_train = self.snpreader_whole[train_idx,:]
covariate_train3 = self.covariate_whole[train_idx,:].read()
covariate_train3.val = np.array([[float(num)] for num in xrange(covariate_train3.iid_count)])
pheno_train3 = self.pheno_whole[train_idx,:].read()
np.random.seed(0)
pheno_train3.val = covariate_train3.val * 2.0 + 100 + np.random.normal(size=covariate_train3.val.shape) # y = 2*x+100+normal(0,1)
##Plot training x and y
#pylab.plot(covariate_train3.val, pheno_train3.val,".")
#pylab.show()
for force_full_rank,force_low_rank in [(True,False),(False,True)]:
#Learn model, save, load
fastlmm3x = FastLMM(force_full_rank=force_full_rank,force_low_rank=force_low_rank,GB_goal=2).fit(K0_train=G0_train, X=covariate_train3, y=pheno_train3)
filename = self.tempout_dir + "/model_lr.flm.p"
pstutil.create_directory_if_necessary(filename)
joblib.dump(fastlmm3x, filename)
fastlmm3 = joblib.load(filename)
#Predict with model (test on train)
predicted_pheno, covar = fastlmm3.predict(K0_whole_test=G0_train, X=covariate_train3,count_A1=False) #test on train
output_file = self.file_name("lr")
Dat.write(output_file,predicted_pheno)
## Plot training x and y, and training x with predicted y
#do_plot = True
#if do_plot:
# pylab.plot(covariate_train3.val, pheno_train3.val,covariate_train3.val,predicted_pheno.val,".")
# pylab.show()
# # Plot y and predicted y (test on train)
# pheno_actual = pheno_train3.val[:,0]
# pylab.plot(pheno_actual,predicted_pheno.val,".")
# pylab.show()
self.compare_files(predicted_pheno,"lr")
def compare_files(self,answer,ref_base):
reffile = TestFeatureSelection.reference_file("fastlmm/"+ref_base+".dat") #Uses same results folder as lmm_train
reference=Dat(reffile).read()
assert np.array_equal(answer.col,reference.col), "sid differs. File '{0}'".format(reffile)
assert np.array_equal(answer.row,reference.row), "iid differs. File '{0}'".format(reffile)
for iid_index in range(reference.row_count):
for sid_index in range(reference.col_count):
a_v = answer.val[iid_index,sid_index]
r_v = reference.val[iid_index,sid_index]
assert abs(a_v - r_v) < 1e-4, "Value at {0},{1} differs too much from file '{2}'".format(iid_index,sid_index,reffile)
def test_kernel(self):
logging.info("TestLmmTrain test_kernel")
train_idx = np.r_[10:self.snpreader_whole.iid_count] # iids 10 and on
test_idx = np.r_[0:10] # the first 10 iids
# Show it using the snps
K0_train = self.snpreader_whole[train_idx,:].read_kernel(Unit())
covariate_train3 = self.covariate_whole[train_idx,:].read()
pheno_train3 = self.pheno_whole[train_idx,:].read()
pheno_train3.val = self.snpreader_whole[train_idx,0:1].read().val*2
assert np.array_equal(K0_train.iid,covariate_train3.iid), "Expect iids to be the same (so that early and late Unit standardization will give the same result)"
assert np.array_equal(K0_train.iid,pheno_train3.iid), "Expect iids to be the same (so that early and late Unit standardization will give the same result)"
#pylab.plot(G0_train[:,0:1].read().val[:,0], pheno_train3.val[:,0],".")
#pylab.show()
#Learn model, save, load
fastlmm3x = FastLMM(GB_goal=2).fit(K0_train=K0_train, X=covariate_train3, y=pheno_train3)
filename = self.tempout_dir + "/model_snps.flm.p"
pstutil.create_directory_if_necessary(filename)
joblib.dump(fastlmm3x, filename)
fastlmm3 = joblib.load(filename)
#Predict with model (test on train)
predicted_pheno, covar = fastlmm3.predict(K0_whole_test=K0_train, X=covariate_train3,count_A1=False) #test on train
output_file = self.file_name("kernel")
Dat.write(output_file,predicted_pheno)
#### Plot training x and y, and training x with predicted y
#pylab.plot(self.snpreader_whole[train_idx,0:1].read().val[:,0], pheno_train3.val,".",self.snpreader_whole[train_idx,0:1].read().val[:,0],predicted_pheno.val,".")
#pylab.show()
#### Plot y and predicted y (test on train)
#pheno_actual = pheno_train3.val[:,0]
#pylab.plot(pheno_actual,predicted_pheno.val,".")
#pylab.show()
self.compare_files(predicted_pheno,"snps") #"kernel" and "snps" test cases should give the same results
def test_str2(self):
logging.info("TestLmmTrain test_str2")
#Standardize train and test together
whole_kernel = self.snpreader_whole.read_kernel(Unit())
train_idx = np.r_[10:self.snpreader_whole.iid_count] # iids 10 and on
test_idx = np.r_[0:10] # the first 10 iids
covariate_train = self.covariate_whole[train_idx,:]
pheno_train = self.pheno_whole[train_idx,:]
K0_train_filename = self.tempout_dir + "/model_str2.kernel.npz"
pstutil.create_directory_if_necessary(K0_train_filename)
from pysnptools.kernelreader import KernelNpz
KernelNpz.write(K0_train_filename,whole_kernel[train_idx].read(order='A',view_ok=True))
fastlmm1 = FastLMM(GB_goal=2).fit(K0_train=K0_train_filename, X=covariate_train, y=pheno_train)
filename = self.tempout_dir + "/model_str2.flm.p"
pstutil.create_directory_if_necessary(filename)
joblib.dump(fastlmm1, filename)
fastlmm2 = joblib.load(filename)
# predict on test set
G0_test = self.snpreader_whole[test_idx,:]
covariate_test = self.covariate_whole[test_idx,:]
predicted_pheno, covar = fastlmm2.predict(K0_whole_test=whole_kernel[:,test_idx].read(order='A',view_ok=True), X=covariate_test,count_A1=False)
output_file = self.file_name("str2")
Dat.write(output_file,predicted_pheno)
#pheno_actual = self.pheno_whole[test_idx,:].read().val[:,0]
#pylab.plot(pheno_actual, predicted_pheno.val,".")
#pylab.show()
self.compare_files(predicted_pheno,"str2")
def test_snps(self):
logging.info("TestLmmTrain test_snps")
train_idx = np.r_[10:self.snpreader_whole.iid_count] # iids 10 and on
test_idx = np.r_[0:10] # the first 10 iids
# Show it using the snps
G0_train = self.snpreader_whole[train_idx,:]
covariate_train3 = self.covariate_whole[train_idx,:].read()
pheno_train3 = self.pheno_whole[train_idx,:].read()
pheno_train3.val = G0_train[:,0:1].read().val*2
#pylab.plot(G0_train[:,0:1].read().val[:,0], pheno_train3.val[:,0],".")
#pylab.show()
#Learn model, save, load
fastlmm3x = FastLMM(GB_goal=2).fit(K0_train=G0_train, X=covariate_train3, y=pheno_train3)
filename = self.tempout_dir + "/model_snps.flm.p"
pstutil.create_directory_if_necessary(filename)
joblib.dump(fastlmm3x, filename)
fastlmm3 = joblib.load(filename)
#Predict with model (test on train)
predicted_pheno, covar = fastlmm3.predict(K0_whole_test=G0_train, X=covariate_train3,count_A1=False) #test on train
output_file = self.file_name("snps")
Dat.write(output_file,predicted_pheno)
### Plot training x and y, and training x with predicted y
#pylab.plot(G0_train[:,0:1].read().val[:,0], pheno_train3.val,".",G0_train[:,0:1].read().val[:,0],predicted_pheno.val,".")
#pylab.show()
### Plot y and predicted y (test on train)
#pheno_actual = pheno_train3.val[:,0]
#pylab.plot(pheno_actual,predicted_pheno.val,".")
#pylab.show()
self.compare_files(predicted_pheno,"snps")
def factory_iterator():
snp_reader_factory_bed = lambda: Bed("examples/toydata",
count_A1=False)
snp_reader_factory_snpmajor_hdf5 = lambda: SnpHdf5(
"examples/toydata.snpmajor.snp.hdf5")
snp_reader_factory_iidmajor_hdf5 = lambda: SnpHdf5(
"examples/toydata.iidmajor.snp.hdf5")
snp_reader_factory_dat = lambda: Dat("examples/toydata.dat")
previous_wd = os.getcwd()
os.chdir(os.path.dirname(os.path.realpath(__file__)))
snpreader0 = snp_reader_factory_bed()
S_original = snpreader0.sid_count
N_original = snpreader0.iid_count
snps_to_read_count = min(S_original, 100)
for iid_index_list in [
list(range(N_original)),
list(range(N_original / 2)),
list(range(N_original - 1, 0, -2))
]:
for snp_index_list in [
list(range(snps_to_read_count)),
list(range(snps_to_read_count / 2)),
list(range(snps_to_read_count - 1, 0, -2))
]:
for standardizer in [Unit(), Beta(1, 25)]:
reference_snps, reference_dtype = NaNCNCTestCases(
iid_index_list, snp_index_list, standardizer,
snp_reader_factory_bed(), sp.float64, "C", "False",
None, None).read_and_standardize()
for snpreader_factory in [
snp_reader_factory_bed,
snp_reader_factory_snpmajor_hdf5,
snp_reader_factory_iidmajor_hdf5,
snp_reader_factory_dat
]:
for dtype in [sp.float64, sp.float32]:
for order in ["C", "F"]:
for force_python_only in [False, True]:
snpreader = snpreader_factory()
test_case = NaNCNCTestCases(
iid_index_list, snp_index_list,
standardizer, snpreader, dtype, order,
force_python_only, reference_snps,
reference_dtype)
yield test_case
os.chdir(previous_wd)
def setUpClass(self):
currentFolder = os.path.dirname(os.path.realpath(__file__))
self.snpreader_bed = Bed(currentFolder + "/examples/toydata")
#Creating Hdf5 data ...
#snpData = self.snpreader_bed.read()
#Hdf5.write(snpData, currentFolder + "/examples/toydata.snpmajor.hdf5")
#Hdf5.write(snpData, currentFolder + "/examples/toydata.iidmajor.hdf5",snp_major=False)
#Creating Dat data ...
#snpData = self.snpreader_bed.read()
#Dat.write(snpData, currentFolder + "/examples/toydata.dat")
###Creating Ped data ...
#snpData = self.snpreader_bed.read()
#Ped.write(snpData, currentFolder + "/examples/toydata.ped")
#fromPed = Ped(currentFolder + "/examples/toydata").read()
#self.assertTrue(np.allclose(snpData.val, fromPed.val, rtol=1e-05, atol=1e-05))
self.snpreader_hdf5 = Hdf5(currentFolder +
"/examples/toydata.snpmajor.hdf5")
self.snpreader_dat = Dat(currentFolder + "/examples/toydata.dat")
self.snpreader_ped = Ped(currentFolder + "/examples/toydata")
self.pheno_fn = currentFolder + "/examples/toydata.phe"
self.pheno_shuffleplus_fn = currentFolder + "/examples/toydata.shufflePlus.phe"
def test_c_reader_dat(self):
snpreader = Dat(self.currentFolder + "/examples/toydata.dat")[:,::100]
_fortesting_JustCheckExists().input(snpreader)
snpdata1 = snpreader.read()
self.assertEqual(np.float64, snpdata1.val.dtype)
self.assertTrue(np.allclose(self.snps[:,::100], snpdata1.val, rtol=1e-05, atol=1e-05))
snpdata1.val[1,2] = np.NaN # Inject a missing value to test writing and reading missing values
output = "tempdir/snpreader/toydata.dat"
create_directory_if_necessary(output)
Dat.write(output,snpdata1)
snpdata2 = Dat(output).read()
np.testing.assert_array_almost_equal(snpdata1.val, snpdata2.val, decimal=10)
snpdata3 = snpdata1[:,0:0].read() #create snpdata with no sids
output = "tempdir/snpreader/toydata3.dat"
Dat.write(output,snpdata3)
snpdata4 = Dat(output).read()
assert snpdata3 == snpdata4
def test_lr_real(self):
do_plot = False
import pylab
logging.info("TestLinRegTrain test_lr_real")
train_idx = np.r_[10:self.snpreader_whole.iid_count] # iids 10 and on
test_idx = np.r_[0:10] # the first 10 iids
#make covar just numbers 0,1,...
covar = self.covariate_whole.read()
covar.val = np.array([[float(num)] for num in xrange(covar.iid_count)])
covariate_train = covar[train_idx, :].read()
covariate_test = covar[test_idx, :].read()
K0_test_test = KernelIdentity(covariate_test.iid)
#make pheno # pheno = 2*covar+100+normal(0,1)*10
pheno = self.pheno_whole.read()
np.random.seed(0)
pheno.val = covar.val * 2.0 + 100 + np.random.normal(
size=covar.val.shape) * 10
pheno_train = pheno[train_idx, :].read()
pheno_test = pheno[test_idx, :].read()
if do_plot:
#Plot training x and y, testing x and y
pylab.plot(covariate_train.val, pheno_train.val, ".",
covariate_test.val, pheno_test.val, ".")
pylab.suptitle("Plot training x and y, testing x and y")
pylab.show()
Xtrain = np.c_[covariate_train.val,
np.ones((covariate_train.iid_count, 1))]
Xtest = np.c_[covariate_test.val,
np.ones((covariate_test.iid_count, 1))]
lsqSol = np.linalg.lstsq(Xtrain, pheno_train.val[:, 0], rcond=-1)
bs = lsqSol[0] #weights
r2 = lsqSol[1] #squared residuals
D = lsqSol[2] #rank of design matrix
N = pheno_train.iid_count
REML = False
if not REML:
sigma2 = float(r2 / N)
nLL = N * 0.5 * np.log(2 * np.pi * sigma2) + N * 0.5
else:
sigma2 = float(r2 / (N - D))
nLL = N * 0.5 * np.log(2 * np.pi * sigma2) + 0.5 / sigma2 * r2
nLL -= 0.5 * D * np.log(2 * np.pi * sigma2)
#REML term
predicted = Xtest.dot(bs)
yerr = [np.sqrt(sigma2)] * len(predicted)
if do_plot:
pylab.plot(covariate_test.val, pheno_test.val, "g.",
covariate_test.val, predicted, "r.")
pylab.xlim([-1, 10])
pylab.errorbar(covariate_test.val,
predicted,
yerr,
linestyle='None')
pylab.suptitle("real linear regression: actual to prediction")
pylab.show()
#These should all give the same result
first_name = None
for name, K0_train, K0_whole_test in [("Identity Kernel", None, None)]:
first_name = first_name or name
#Learn model, save, load
modelx = LinearRegression().fit(K0_train=K0_train,
X=covariate_train,
y=pheno_train)
filename = self.tempout_dir + "/model_lr_real.flm.p"
pstutil.create_directory_if_necessary(filename)
joblib.dump(modelx, filename)
model = joblib.load(filename)
do_test_on_train = True
if do_test_on_train:
#Predict with model (test on train)
predicted_pheno, covar = model.predict(
K0_whole_test=K0_train, X=covariate_train) #test on train
output_file = self.file_name("lr_reala_" + name)
Dat.write(output_file, predicted_pheno)
covar2 = SnpData(
iid=covar.row, sid=covar.col[:, 1],
val=covar.val) #kludge to write kernel to text format
output_file = self.file_name("lr_reala.cov_" + name)
Dat.write(output_file, covar2)
yerr = np.sqrt(np.diag(covar.val))
predicted = predicted_pheno.val
if do_plot:
pylab.plot(covariate_train.val, pheno_train.val, "g.",
covariate_train.val, predicted, "r.")
pylab.xlim([0, 50])
pylab.ylim([100, 200])
pylab.errorbar(covariate_train.val,
predicted,
yerr,
linestyle='None')
pylab.suptitle(
name +
": test on train: train X to true target (green) and prediction (red)"
)
pylab.show()
self.compare_files(predicted_pheno, "lr2a_" + first_name)
self.compare_files(covar2, "lr2a.cov_" + first_name)
#Predict with model (test on test)
predicted_pheno, covar = model.predict(
K0_whole_test=K0_whole_test, X=covariate_test) #test on train
output_file = self.file_name("lr_realb_" + name)
Dat.write(output_file, predicted_pheno)
covar2 = SnpData(
iid=covar.row, sid=covar.col[:, 1],
val=covar.val) #kludge to write kernel to text format
output_file = self.file_name("lr_realb.cov_" + name)
Dat.write(output_file, covar2)
yerr = np.sqrt(np.diag(covar.val))
predicted = predicted_pheno.val
if do_plot:
pylab.plot(covariate_test.val, pheno_test.val, "g.",
covariate_test.val, predicted, "r.")
pylab.xlim([-1, 10])
pylab.errorbar(covariate_test.val,
predicted,
yerr,
linestyle='None')
pylab.suptitle(
name +
": test on test: test X to true target (green) and prediction (red)"
)
pylab.show()
## Plot y and predicted y (test on train)
#pylab.plot(pheno_test.val,predicted_pheno.val,".")
#pylab.suptitle(name+": test on test: true target to prediction")
#pylab.show()
self.compare_files(predicted_pheno, "lr2b_" + first_name)
self.compare_files(covar2, "lr2b.cov_" + first_name)
def test_lr_as_lmm(self):
do_plot = False
#later why does this test case generate two intersect info messages instead of just one?
import pylab
logging.info("TestLmmTrain test_lr_as_lmm")
###############################################################
# Create a linear data set with just a little noise
###############################################################
train_idx = np.r_[10:self.snpreader_whole.iid_count] # iids 10 and on
test_idx = np.r_[0:10] # the first 10 iids
#make covar just numbers 0,1,...
covar = self.covariate_whole.read()
covar.val = np.array([[float(num)] for num in xrange(covar.iid_count)])
covar._name = 'np.array([[float(num)] for num in xrange(covar.iid_count)])'
covariate_train = covar[train_idx,:].read()
covariate_test = covar[test_idx,:].read()
#make pheno # pheno = 2*covar+100+normal(0,1)*10
pheno = self.pheno_whole.read()
np.random.seed(0)
pheno.val = covar.val * 2.0 + 100 + np.random.normal(size=covar.val.shape)*10
pheno_train = pheno[train_idx,:].read()
pheno_test = pheno[test_idx,:].read()
if do_plot:
#Plot training x and y, testing x and y
pylab.plot(covariate_train.val, pheno_train.val,".",covariate_test.val, pheno_test.val,".")
pylab.suptitle("Plot training x and y, testing x and y")
pylab.show()
###############################################################
# Show that linear regression does a good job predicting
###############################################################
Xtrain = np.c_[covariate_train.val,np.ones((covariate_train.iid_count,1))]
Xtest = np.c_[covariate_test.val,np.ones((covariate_test.iid_count,1))]
lsqSol = np.linalg.lstsq(Xtrain, pheno_train.val[:,0],rcond=-1)
bs=lsqSol[0] #weights
r2=lsqSol[1] #squared residuals
D=lsqSol[2] #rank of design matrix
N=pheno_train.iid_count
REML = False
if not REML:
sigma2 = float(r2/N)
nLL = N*0.5*np.log(2*np.pi*sigma2) + N*0.5
else:
sigma2 = float(r2 / (N-D))
nLL = N*0.5*np.log(2*np.pi*sigma2) + 0.5/sigma2*r2;
nLL -= 0.5*D*np.log(2*np.pi*sigma2);#REML term
predicted = Xtest.dot(bs)
yerr = [np.sqrt(sigma2)] * len(predicted)
if do_plot:
pylab.plot(covariate_test.val, pheno_test.val,"g.",covariate_test.val, predicted,"r.")
pylab.xlim([-1, 10])
pylab.errorbar(covariate_test.val, predicted,yerr,linestyle='None')
pylab.suptitle("real linear regression: actual to prediction")
pylab.show()
###############################################################
# Use LMM as LR and apply test on train
###############################################################
for force_full_rank in [True, False]:
#Learn model, save, load
fastlmmx = FastLMM(GB_goal=2,force_full_rank=force_full_rank).fit(K0_train=covariate_train, X=None, y=pheno_train)
filename = self.tempout_dir + "/model_lr_as_lmm.flm.p"
pstutil.create_directory_if_necessary(filename)
joblib.dump(fastlmmx, filename)
fastlmm = joblib.load(filename)
do_test_on_train = True
if do_test_on_train:
#Predict with model (test on train)
predicted_pheno, covar = fastlmm.predict(K0_whole_test=covariate_train, X=None,count_A1=False) #test on train
output_file = self.file_name("lr_as_lmma_")
Dat.write(output_file,predicted_pheno)
covar2 = SnpData(iid=covar.row,sid=covar.col[:,1],val=covar.val) #kludge to write kernel to text format
output_file = self.file_name("lr_as_lmma.cov_")
Dat.write(output_file,covar2)
yerr = np.sqrt(np.diag(covar.val))
predicted = predicted_pheno.val
if do_plot:
pylab.plot(covariate_train.val, pheno_train.val,"g.",covariate_train.val, predicted,"r.")
pylab.xlim([0, 50])
pylab.ylim([100, 200])
pylab.errorbar(covariate_train.val, predicted,yerr,linestyle='None')
pylab.suptitle("test on train: train X to true target (green) and prediction (red)")
pylab.show()
self.compare_files(predicted_pheno,"lr_as_lmma_")
self.compare_files(covar2,"lr_as_lmma.cov_")
###############################################################
# Use LMM as LR and apply test on test
###############################################################
#Predict with model (test on test)
predicted_pheno, covar = fastlmm.predict(K0_whole_test=covariate_test, X=None,count_A1=False) #test on train
output_file = self.file_name("lr_as_lmmb_")
Dat.write(output_file,predicted_pheno)
covar2 = SnpData(iid=covar.row,sid=covar.col[:,1],val=covar.val) #kludge to write kernel to text format
output_file = self.file_name("lr_as_lmmb.cov_")
Dat.write(output_file,covar2)
yerr = np.sqrt(np.diag(covar.val))
predicted = predicted_pheno.val
if do_plot:
pylab.plot(covariate_test.val, pheno_test.val,"g.",covariate_test.val, predicted,"r.")
pylab.xlim([-1, 10])
pylab.errorbar(covariate_test.val, predicted,yerr,linestyle='None')
pylab.suptitle("test on test: test X to true target (green) and prediction (red)")
pylab.show()
## Plot y and predicted y (test on train)
#pylab.plot(pheno_test.val,predicted_pheno.val,".")
#pylab.suptitle(name+": test on test: true target to prediction")
#pylab.show()
self.compare_files(predicted_pheno,"lr_as_lmmb_")
self.compare_files(covar2,"lr_as_lmmb.cov_")
def test_lr_real(self):
do_plot = False
import pylab
logging.info("TestLinRegTrain test_lr_real")
train_idx = np.r_[10:self.snpreader_whole.iid_count] # iids 10 and on
test_idx = np.r_[0:10] # the first 10 iids
#make covar just numbers 0,1,...
covar = self.covariate_whole.read()
covar.val = np.array([[float(num)] for num in xrange(covar.iid_count)])
covariate_train = covar[train_idx,:].read()
covariate_test = covar[test_idx,:].read()
K0_test_test = KernelIdentity(covariate_test.iid)
#make pheno # pheno = 2*covar+100+normal(0,1)*10
pheno = self.pheno_whole.read()
np.random.seed(0)
pheno.val = covar.val * 2.0 + 100 + np.random.normal(size=covar.val.shape)*10
pheno_train = pheno[train_idx,:].read()
pheno_test = pheno[test_idx,:].read()
if do_plot:
#Plot training x and y, testing x and y
pylab.plot(covariate_train.val, pheno_train.val,".",covariate_test.val, pheno_test.val,".")
pylab.suptitle("Plot training x and y, testing x and y")
pylab.show()
Xtrain = np.c_[covariate_train.val,np.ones((covariate_train.iid_count,1))]
Xtest = np.c_[covariate_test.val,np.ones((covariate_test.iid_count,1))]
lsqSol = np.linalg.lstsq(Xtrain, pheno_train.val[:,0])
bs=lsqSol[0] #weights
r2=lsqSol[1] #squared residuals
D=lsqSol[2] #rank of design matrix
N=pheno_train.iid_count
REML = False
if not REML:
sigma2 = float(r2/N)
nLL = N*0.5*np.log(2*np.pi*sigma2) + N*0.5
else:
sigma2 = float(r2 / (N-D))
nLL = N*0.5*np.log(2*np.pi*sigma2) + 0.5/sigma2*r2;
nLL -= 0.5*D*np.log(2*np.pi*sigma2);#REML term
predicted = Xtest.dot(bs)
yerr = [np.sqrt(sigma2)] * len(predicted)
if do_plot:
pylab.plot(covariate_test.val, pheno_test.val,"g.",covariate_test.val, predicted,"r.")
pylab.xlim([-1, 10])
pylab.errorbar(covariate_test.val, predicted,yerr,linestyle='None')
pylab.suptitle("real linear regression: actual to prediction")
pylab.show()
#These should all give the same result
first_name = None
for name,K0_train,K0_whole_test in [("Identity Kernel",None,None)]:
first_name = first_name or name
#Learn model, save, load
modelx = LinearRegression().fit(K0_train=K0_train, X=covariate_train, y=pheno_train)
filename = self.tempout_dir + "/model_lr_real.flm.p"
pstutil.create_directory_if_necessary(filename)
joblib.dump(modelx, filename)
model = joblib.load(filename)
do_test_on_train = True
if do_test_on_train:
#Predict with model (test on train)
predicted_pheno, covar = model.predict(K0_whole_test=K0_train, X=covariate_train) #test on train
output_file = self.file_name("lr_reala_"+name)
Dat.write(output_file,predicted_pheno)
covar2 = SnpData(iid=covar.row,sid=covar.col[:,1],val=covar.val) #kludge to write kernel to text format
output_file = self.file_name("lr_reala.cov_"+name)
Dat.write(output_file,covar2)
yerr = np.sqrt(np.diag(covar.val))
predicted = predicted_pheno.val
if do_plot:
pylab.plot(covariate_train.val, pheno_train.val,"g.",covariate_train.val, predicted,"r.")
pylab.xlim([0, 50])
pylab.ylim([100, 200])
pylab.errorbar(covariate_train.val, predicted,yerr,linestyle='None')
pylab.suptitle(name+": test on train: train X to true target (green) and prediction (red)")
pylab.show()
self.compare_files(predicted_pheno,"lr2a_"+first_name)
self.compare_files(covar2,"lr2a.cov_"+first_name)
#Predict with model (test on test)
predicted_pheno, covar = model.predict(K0_whole_test=K0_whole_test, X=covariate_test) #test on train
output_file = self.file_name("lr_realb_"+name)
Dat.write(output_file,predicted_pheno)
covar2 = SnpData(iid=covar.row,sid=covar.col[:,1],val=covar.val) #kludge to write kernel to text format
output_file = self.file_name("lr_realb.cov_"+name)
Dat.write(output_file,covar2)
yerr = np.sqrt(np.diag(covar.val))
predicted = predicted_pheno.val
if do_plot:
pylab.plot(covariate_test.val, pheno_test.val,"g.",covariate_test.val, predicted,"r.")
pylab.xlim([-1, 10])
pylab.errorbar(covariate_test.val, predicted,yerr,linestyle='None')
pylab.suptitle(name+": test on test: test X to true target (green) and prediction (red)")
pylab.show()
## Plot y and predicted y (test on train)
#pylab.plot(pheno_test.val,predicted_pheno.val,".")
#pylab.suptitle(name+": test on test: true target to prediction")
#pylab.show()
self.compare_files(predicted_pheno,"lr2b_"+first_name)
self.compare_files(covar2,"lr2b.cov_"+first_name)
def test_load_and_standardize_dat(self):
snpreader2 = Dat(self.currentFolder + "/examples/toydata.dat")
self.load_and_standardize(snpreader2, snpreader2)
def test_standardize_dat(self):
snpreader = Dat(self.currentFolder + "/examples/toydata.dat")
self.standardize(snpreader)
def test_lmm(self):
do_plot = False
iid_count = 500
seed = 0
import pylab
logging.info("TestLmmTrain test_lmm")
iid = [["cid{0}P{1}".format(iid_index,iid_index//250)]*2 for iid_index in xrange(iid_count)]
train_idx = np.r_[10:iid_count] # iids 10 and on
test_idx = np.r_[0:10] # the first 10 iids
#Every person is 100% related to everyone in one of 5 families
K0a = KernelData(iid=iid,val=np.empty([iid_count,iid_count]),name="related by distance")
for iid_index0 in xrange(iid_count):
for iid_index1 in xrange(iid_count):
K0a.val[iid_index0,iid_index1] = 1 if iid_index0 % 5 == iid_index1 % 5 else 0
if iid_index1 < iid_index0:
assert K0a.val[iid_index0,iid_index1] == K0a.val[iid_index1,iid_index0]
#every person lives on a line from 0 to 1
# They are related to every other person as a function of distance on the line
np.random.seed(seed)
home = np.random.random([iid_count])
K0b = KernelData(iid=iid,val=np.empty([iid_count,iid_count]),name="related by distance")
for iid_index in xrange(iid_count):
K0b.val[iid_index,:] = 1 - np.abs(home-home[iid_index])**.1
#make covar just numbers 0,1,...
covar = SnpData(iid=iid,sid=["x"],val=np.array([[float(num)] for num in xrange(iid_count)]))
covariate_train = covar[train_idx,:].read()
covariate_test = covar[test_idx,:].read()
for name, h2, K0 in [("clones", 1, K0a),("line_world",.75,K0b)]:
sigma2x = 100
varg = sigma2x * h2
vare = sigma2x * (1-h2)
#######################################################################
#make pheno # pheno = 2*covar+100+normal(0,1)*2.5+normal(0,K)*7.5
#######################################################################
#random.multivariate_normal is sensitive to mkl_num_thread, so we control it.
if 'MKL_NUM_THREADS' in os.environ:
mkl_num_thread = os.environ['MKL_NUM_THREADS']
else:
mkl_num_thread = None
os.environ['MKL_NUM_THREADS'] = '1'
np.random.seed(seed)
p1 = covar.val * 2.0 + 100
p2 = np.random.normal(size=covar.val.shape)*np.sqrt(vare)
p3 = (np.random.multivariate_normal(np.zeros(iid_count),K0.val)*np.sqrt(varg)).reshape(-1,1)
if mkl_num_thread is not None:
os.environ['MKL_NUM_THREADS'] = mkl_num_thread
else:
del os.environ['MKL_NUM_THREADS']
pheno = SnpData(iid=iid,sid=["pheno0"],val= p1 + p2 + p3)
pheno_train = pheno[train_idx,:].read()
pheno_test = pheno[test_idx,:].read()
if do_plot:
#Plot training x and y, testing x and y
pylab.plot(covariate_train.val, pheno_train.val,".",covariate_test.val, pheno_test.val,".")
pylab.suptitle(name + ": Plot training x and y, testing x and y")
pylab.show()
Xtrain = np.c_[covariate_train.val,np.ones((covariate_train.iid_count,1))]
Xtest = np.c_[covariate_test.val,np.ones((covariate_test.iid_count,1))]
lsqSol = np.linalg.lstsq(Xtrain, pheno_train.val[:,0],rcond=-1)
bs=lsqSol[0] #weights
r2=lsqSol[1] #squared residuals
D=lsqSol[2] #rank of design matrix
N=pheno_train.iid_count
REML = False
if not REML:
sigma2 = float(r2/N)
nLL = N*0.5*np.log(2*np.pi*sigma2) + N*0.5
else:
sigma2 = float(r2 / (N-D))
nLL = N*0.5*np.log(2*np.pi*sigma2) + 0.5/sigma2*r2;
nLL -= 0.5*D*np.log(2*np.pi*sigma2);#REML term
predicted = Xtest.dot(bs)
yerr = [np.sqrt(sigma2)] * len(predicted)
if do_plot:
pylab.plot(covariate_test.val, pheno_test.val,"g.",covariate_test.val, predicted,"r.")
pylab.xlim([-1, 10])
pylab.errorbar(covariate_test.val, predicted,yerr,linestyle='None')
pylab.suptitle(name + ": real linear regression: actual to prediction")
pylab.show()
for factor in [1,100,.02]:
K0 = K0.read()
K0.val *= factor
K0_train = K0[train_idx]
K0_whole_test = K0[:,test_idx]
#Learn model, save, load
fastlmmx = FastLMM(GB_goal=2).fit(K0_train=K0_train, X=covariate_train, y=pheno_train)
v2 = np.var(p2)
v3 = np.var(p3)
logging.debug("Original h2 of {0}. Generated h2 of {1}. Learned h2 of {2}".format(h2, v3/(v2+v3), fastlmmx.h2raw))
filename = self.tempout_dir + "/model_lmm.flm.p"
pstutil.create_directory_if_necessary(filename)
joblib.dump(fastlmmx, filename)
fastlmm = joblib.load(filename)
do_test_on_train = True
if do_test_on_train:
#Predict with model (test on train)
predicted_pheno, covar_pheno = fastlmm.predict(K0_whole_test=K0_train, X=covariate_train,count_A1=False) #test on train
output_file = self.file_name("lmma_"+name)
Dat.write(output_file,predicted_pheno)
covar2 = SnpData(iid=covar_pheno.row,sid=covar_pheno.col[:,1],val=covar_pheno.val) #kludge to write kernel to text format
output_file = self.file_name("lmma.cov_"+name)
Dat.write(output_file,covar2)
yerr = np.sqrt(np.diag(covar_pheno.val))
predicted = predicted_pheno.val
if do_plot:
pylab.plot(covariate_train.val, pheno_train.val,"g.",covariate_train.val, predicted,"r.")
pylab.xlim([0, 50])
pylab.ylim([100, 200])
pylab.errorbar(covariate_train.val, predicted,yerr,linestyle='None')
pylab.suptitle(name+": test on train: train X to true target (green) and prediction (red)")
pylab.show()
self.compare_files(predicted_pheno,"lmma_"+name)
self.compare_files(covar2,"lmma.cov_"+name)
predicted_pheno0, covar_pheno0 = fastlmm.predict(K0_whole_test=K0_train[:,0], X=covariate_train[0,:],count_A1=False) #test on train #0
assert np.abs(predicted_pheno0.val[0,0] - predicted_pheno.val[0,0]) < 1e-6, "Expect a single case to get the same prediction as a set of cases"
assert np.abs(covar_pheno0.val[0,0] - covar_pheno.val[0,0]) < 1e-6, "Expect a single case to get the same prediction as a set of cases"
#Predict with model (test on test)
predicted_phenoB, covar_phenoB = fastlmm.predict(K0_whole_test=K0_whole_test, X=covariate_test,count_A1=False) #test on test
output_file = self.file_name("lmmb_"+name)
Dat.write(output_file,predicted_phenoB)
covar2 = SnpData(iid=covar_phenoB.row,sid=covar_phenoB.col[:,1],val=covar_phenoB.val) #kludge to write kernel to text format
output_file = self.file_name("lmmb.cov_"+name)
Dat.write(output_file,covar2)
yerr = np.sqrt(np.diag(covar_phenoB.val))
predicted = predicted_phenoB.val
if do_plot:
pylab.plot(covariate_test.val, pheno_test.val,"g.",covariate_test.val, predicted,"r.")
pylab.xlim([-1, 10])
pylab.errorbar(covariate_test.val, predicted,yerr,linestyle='None')
pylab.suptitle(name+": test on test: test X to true target (green) and prediction (red)")
pylab.show()
self.compare_files(predicted_phenoB,"lmmb_"+name)
self.compare_files(covar2,"lmmb.cov_"+name)
predicted_phenoB0, covar_phenoB0 = fastlmm.predict(K0_whole_test=K0_whole_test[:,0], X=covariate_test[0,:],count_A1=False) #test on a single test case
assert np.abs(predicted_phenoB0.val[0,0] - predicted_phenoB.val[0,0]) < 1e-6, "Expect a single case to get the same prediction as a set of cases"
assert np.abs(covar_phenoB0.val[0,0] - covar_phenoB.val[0,0]) < 1e-6, "Expect a single case to get the same prediction as a set of cases"
#Predict with model test on some train and some test
some_idx = range(covar.iid_count)
some_idx.remove(train_idx[0])
some_idx.remove(test_idx[0])
covariate_some = covar[some_idx,:]
K0_whole_some = K0[:,some_idx]
predicted_phenoC, covar_phenoC = fastlmm.predict(K0_whole_test=K0_whole_some, X=covariate_some,count_A1=False)
for idxC, iidC in enumerate(predicted_phenoC.iid):
meanC = predicted_phenoC.val[idxC]
varC = covar_phenoC.val[idxC,idxC]
if iidC in predicted_pheno.iid:
predicted_pheno_ref = predicted_pheno
covar_pheno_ref = covar_pheno
else:
assert iidC in predicted_phenoB.iid
predicted_pheno_ref = predicted_phenoB
covar_pheno_ref = covar_phenoB
idx_ref = predicted_pheno_ref.iid_to_index([iidC])[0]
mean_ref = predicted_pheno_ref.val[idx_ref]
var_ref = covar_pheno_ref.val[idx_ref,idx_ref]
assert np.abs(meanC - mean_ref) < 1e-6
assert np.abs(varC - var_ref) < 1e-6