class SparseTestCase(unittest.TestCase):
def setUp(self):
self.file_bed = ["./utest/test1_divided_merged.bed", "./utest/test1_different.bed"]
self.file_len = "./utest/test1_divided.len"
self.file_test = "./utest/test1_similar.bed"
self.file_dhs = "./utest/test1_dhs.bed"
self.estimate_extracted = [
"./utest/test1_divided_merged_extracted.vec",
"./utest/test1_different_extracted.vec",
]
self.bsm = BedNMF(self.file_bed, self.file_len, self.file_dhs, 1)
def testDhsExtraction(self):
self.bsm.init_dhs_extraction()
cnt = 0
for efile in self.estimate_extracted:
with open(efile) as ef:
vector = read_vector(ef)
result = ravel(self.bsm.sparse[:, cnt].todense())
print "estimate extracted", vector
print "result extracted", result
self.assertTrue(alltrue(vector == result))
cnt += 1
def testNmfClassification(self):
self.bsm.init_dhs_extraction()
self.bsm.init_factorization()
print self.bsm.classification(self.file_bed[0])
print self.bsm.classification(self.file_bed[1])
print self.bsm.classification(self.file_test)
class CtcfTestCase(unittest.TestCase):
def setUp(self):
self.train_path = glob.glob("./CTCF_train/*.bed")
print self.train_path
self.len_path = "./static_data/hg19_len"
self.dhs_path = "./static_data/DHS_hg19.bed"
self.test_path = "./CTCF_test/5011_peaks.bed"
self.test2_path = "./CTCF_train/1252_peaks.bed"
self.dump_path = "./persist/train.p"
self.bin_size = 50000
self.InitNMF()
def InitNMF(self):
try:
self.bsm = cPickle.load(open(self.dump_path,"r"))
print "Load from database"
except:
print "Init database"
self.bsm = BedNMF(self.train_path, self.len_path, self.dhs_path, self.bin_size)
self.bsm.init_dhs_extraction()
self.bsm.init_factorization(rank = 10)
cPickle.dump(self.bsm, open(self.dump_path, 'w'))
def gen_testClassification(self, test_path):
result = self.bsm.classification(test_path)
pprint(n_nearest(result, self.train_path))
return result
def testClassification1(self):
self.gen_testClassification(self.test_path)
self.gen_testClassification(self.test2_path)
def InitNMF(self):
try:
self.bsm = cPickle.load(open(self.dump_path,"r"))
print "Load from database"
except:
print "Init database"
self.bsm = BedNMF(self.train_path, self.len_path, self.dhs_path, self.bin_size)
self.bsm.init_dhs_extraction()
self.bsm.init_factorization(rank = 10)
cPickle.dump(self.bsm, open(self.dump_path, 'w'))
def setUp(self):
self.file_bed = ["./utest/test1_divided_merged.bed", "./utest/test1_different.bed"]
self.file_len = "./utest/test1_divided.len"
self.file_test = "./utest/test1_similar.bed"
self.file_dhs = "./utest/test1_dhs.bed"
self.estimate_extracted = [
"./utest/test1_divided_merged_extracted.vec",
"./utest/test1_different_extracted.vec",
]
self.bsm = BedNMF(self.file_bed, self.file_len, self.file_dhs, 1)
