def setUp(self):
self.test_data_root = TEST_INVARIANTS['test_data_root']
self.test_write_root = TEST_INVARIANTS['test_write_root']
self.testdb = 'DB1'
self.test_dtname = 'Test1'
self.dbm = DBManager(self.test_write_root)
# ssdata.dsv
self.ssdata_dsv_path = os.path.abspath(
os.path.join(self.test_data_root, 'ssdata.dsv'))
self.ssdata_comment = '#'
ssdata_dsv_fh = DSV.getHandle(self.ssdata_dsv_path)
self.ssdata_dsv1 = DSV(self.dbm,
self.testdb,
ssdata_dsv_fh,
dtname=self.test_dtname,
comment=self.ssdata_comment)
self.ssdata_dsv1.create()
self.ssdata_dsv1.loadAll()
self.ssdata_dsv1.close()
self.ssdata_samples = ('S1', 'S2', 'S3', 'S4', 'S5')
# hierarchy tests
self.pkc2id = {'PKC1': ('R1', 'R2', 'R3', 'R4'), 'PKC2': ('R5', )}
self.pkc = ('PKC1', 'PKC2')
self.size_thr1 = 2
self.cat1 = SubsetSizeCategorizer(self.size_thr1, ID='Cat1')
self.cat1_uniq_le = self.cat1.uniquifyCategory(
self.cat1.ROW_SIZE_LESSER)
self.cat1_uniq_gt = self.cat1.uniquifyCategory(
self.cat1.ROW_SIZE_GREATER)
self.size_thr2 = 3
self.cat2 = SubsetSizeCategorizer(self.size_thr2, ID='Cat2')
self.cat2_uniq_le = self.cat2.uniquifyCategory(
self.cat2.ROW_SIZE_LESSER)
self.cat2_uniq_gt = self.cat2.uniquifyCategory(
self.cat2.ROW_SIZE_GREATER)
self.size_thr3 = 0
self.cat3 = SubsetSizeCategorizer(self.size_thr3, ID='Cat3')
self.cat3_uniq_le = self.cat3.uniquifyCategory(
self.cat3.ROW_SIZE_LESSER)
self.cat3_uniq_gt = self.cat3.uniquifyCategory(
self.cat3.ROW_SIZE_GREATER)
self.cinst = {
'Cat1': self.cat1,
'Cat2': self.cat2,
'Cat3': self.cat3,
}
self.cmap1 = ['Cat1', 'Cat2', 'Cat3']
self.cmap2 = ['Cat3', 'Cat1', 'Cat2']
self.symbols = list(self.pkc)
def setUp(self):
self.test_data_root = TEST_INVARIANTS['test_data_root']
self.test_write_root = TEST_INVARIANTS['test_write_root']
self.testdb = 'DB1'
self.test_dtname = 'Test1'
self.dbm = DBManager(self.test_write_root)
# ssdata.dsv
self.ssdata_dsv_path = os.path.abspath(
os.path.join(self.test_data_root, 'ssdata.dsv'))
self.ssdata_comment = '#'
ssdata_dsv_fh = DSV.getHandle(self.ssdata_dsv_path)
self.ssdata_dsv1 = DSV(self.dbm,
self.testdb,
ssdata_dsv_fh,
dtname=self.test_dtname,
comment=self.ssdata_comment)
self.ssdata_dsv1.create()
self.ssdata_dsv1.loadAll()
self.ssdata_dsv1.close()
self.ssdata_samples = ('S1', 'S2', 'S3', 'S4', 'S5')
# pkcidmap
self.pkc2id1 = {
'PKC1': ('R1', 'R2'),
'PKC2': ('R3', ),
'PKC3': ('R4', 'R5')
}
self.pkc1 = ('PKC1', 'PKC2', 'PKC3')
self.ss_cols1 = '*'
self.ref_ss1 = [
({
'pkcID': 'PKC1',
'dtable': self.ssdata_dsv1,
'rows': ['R1', 'R2'],
'cols': list(self.ssdata_samples)
}, None),
({
'pkcID': 'PKC2',
'dtable': self.ssdata_dsv1,
'rows': ['R3'],
'cols': list(self.ssdata_samples)
}, None),
({
'pkcID': 'PKC3',
'dtable': self.ssdata_dsv1,
'rows': ['R4', 'R5'],
'cols': list(self.ssdata_samples)
}, None),
]
self.ss_cols2 = ('S1', 'S4', 'S5')
self.ref_ss2 = [
({
'pkcID': 'PKC1',
'dtable': self.ssdata_dsv1,
'rows': ['R1', 'R2'],
'cols': list(self.ss_cols2)
}, None),
({
'pkcID': 'PKC2',
'dtable': self.ssdata_dsv1,
'rows': ['R3'],
'cols': list(self.ss_cols2)
}, None),
({
'pkcID': 'PKC3',
'dtable': self.ssdata_dsv1,
'rows': ['R4', 'R5'],
'cols': list(self.ss_cols2)
}, None),
]
self.pkc2 = ('PKC3', 'PKC1')
self.ref_ss3 = [
({
'pkcID': 'PKC3',
'dtable': self.ssdata_dsv1,
'rows': ['R4', 'R5'],
'cols': list(self.ss_cols2)
}, None),
({
'pkcID': 'PKC1',
'dtable': self.ssdata_dsv1,
'rows': ['R1', 'R2'],
'cols': list(self.ss_cols2)
}, None),
]
self.ref_ss4 = [
(None,
numpy.array([
[2.44753543273, 42.9497086717, 30.8331998765],
[42.1888598933, 39.1743921225, 15.9744094108],
])),
(None,
numpy.array([
[16.5734780715, 14.8233987496, 21.7385342744],
[60.0958378228, 98.4321570519, 71.9193619126],
])),
]
# for categorization tests
self.pkc2id2 = {'PKC1': ('R1', 'R2', 'R3', 'R4'), 'PKC2': ('R5', )}
self.pkc3 = ('PKC1', 'PKC2')
self.size_thr = 3
self.cat1 = SubsetSizeCategorizer(self.size_thr)
self.exp_categories1 = ['>', '<=']
self.cat2 = NullCategorizer()
self.exp_categories2 = [self.cat2.NULL] * len(self.pkc3)
def test_categorize4(self):
sc = SubsetSizeCategorizer(self.size_thr)
categories = [sc.categorize(ds) for ds in self.ds4]
self.assertSequenceEqual(self.exp_cat4, categories)
class TestPKDrivenDBSubsetHierarchy1(unittest.TestCase):
def setUp(self):
self.test_data_root = TEST_INVARIANTS['test_data_root']
self.test_write_root = TEST_INVARIANTS['test_write_root']
self.testdb = 'DB1'
self.test_dtname = 'Test1'
self.dbm = DBManager(self.test_write_root)
# ssdata.dsv
self.ssdata_dsv_path = os.path.abspath(
os.path.join(self.test_data_root, 'ssdata.dsv'))
self.ssdata_comment = '#'
ssdata_dsv_fh = DSV.getHandle(self.ssdata_dsv_path)
self.ssdata_dsv1 = DSV(self.dbm,
self.testdb,
ssdata_dsv_fh,
dtname=self.test_dtname,
comment=self.ssdata_comment)
self.ssdata_dsv1.create()
self.ssdata_dsv1.loadAll()
self.ssdata_dsv1.close()
self.ssdata_samples = ('S1', 'S2', 'S3', 'S4', 'S5')
# hierarchy tests
self.pkc2id = {'PKC1': ('R1', 'R2', 'R3', 'R4'), 'PKC2': ('R5', )}
self.pkc = ('PKC1', 'PKC2')
self.size_thr1 = 2
self.cat1 = SubsetSizeCategorizer(self.size_thr1, ID='Cat1')
self.cat1_uniq_le = self.cat1.uniquifyCategory(
self.cat1.ROW_SIZE_LESSER)
self.cat1_uniq_gt = self.cat1.uniquifyCategory(
self.cat1.ROW_SIZE_GREATER)
self.size_thr2 = 3
self.cat2 = SubsetSizeCategorizer(self.size_thr2, ID='Cat2')
self.cat2_uniq_le = self.cat2.uniquifyCategory(
self.cat2.ROW_SIZE_LESSER)
self.cat2_uniq_gt = self.cat2.uniquifyCategory(
self.cat2.ROW_SIZE_GREATER)
self.size_thr3 = 0
self.cat3 = SubsetSizeCategorizer(self.size_thr3, ID='Cat3')
self.cat3_uniq_le = self.cat3.uniquifyCategory(
self.cat3.ROW_SIZE_LESSER)
self.cat3_uniq_gt = self.cat3.uniquifyCategory(
self.cat3.ROW_SIZE_GREATER)
self.cinst = {
'Cat1': self.cat1,
'Cat2': self.cat2,
'Cat3': self.cat3,
}
self.cmap1 = ['Cat1', 'Cat2', 'Cat3']
self.cmap2 = ['Cat3', 'Cat1', 'Cat2']
self.symbols = list(self.pkc)
def tearDown(self):
self.dbm.close()
db1_path = os.path.abspath('%s/%s.db' %
(self.test_write_root, self.testdb))
rootdb_path = os.path.abspath('%s/%s.root.db' %
(self.test_write_root, SYSTEM_NAME_LC))
if os.path.exists(db1_path):
os.remove(db1_path)
if os.path.exists(rootdb_path):
os.remove(rootdb_path)
self.dbm = None
def test_init1(self):
pkdm = PKDrivenDBDataManager(self.ssdata_dsv1,
MockPKCIDMap(self.pkc2id))
pkdss = PKDrivenDBSubsetHierarchy(pkdm, self.ssdata_samples)
pkdss.build(self.cmap1, self.cinst, self.symbols)
# ['Cat1', 'Cat2', 'Cat3']
expected_hierarchy = dict()
expected_hierarchy.update({None: self.cat1.id})
expected_hierarchy.update({self.cat1_uniq_le: self.cat2.id})
expected_hierarchy.update({self.cat1_uniq_gt: self.cat2.id})
expected_hierarchy.update({self.cat2_uniq_le: self.cat3.id})
expected_hierarchy.update({self.cat2_uniq_gt: self.cat3.id})
expected_hierarchy.update({self.cat3_uniq_gt: None})
expected_hierarchy.update({self.cat3_uniq_le: None})
self.assertEqual(expected_hierarchy, pkdss.hierarchy)
# [PKC1 PKC2]
# Cat1 [PKC1]>2 [PKC2]<=2
# Cat2 [PKC1]>3 [PKC2]<=3
# Cat3 [PKC1]>0 [PKC2]>0
expected_symboltree = dict()
expected_symboltree.update(
{None: {
self.cat1_uniq_gt: ['PKC1'],
self.cat1_uniq_le: ['PKC2']
}})
expected_symboltree.update(
{self.cat1_uniq_gt: {
self.cat2_uniq_gt: ['PKC1']
}})
expected_symboltree.update(
{self.cat1_uniq_le: {
self.cat2_uniq_le: ['PKC2']
}})
expected_symboltree.update(
{self.cat2_uniq_gt: {
self.cat3_uniq_gt: ['PKC1']
}})
expected_symboltree.update(
{self.cat2_uniq_le: {
self.cat3_uniq_gt: ['PKC2']
}})
self.assertEqual(expected_symboltree, pkdss.symboltree)
def test_init2(self):
pkdm = PKDrivenDBDataManager(self.ssdata_dsv1,
MockPKCIDMap(self.pkc2id))
pkdss = PKDrivenDBSubsetHierarchy(pkdm, self.ssdata_samples)
pkdss.build(self.cmap2, self.cinst, self.symbols)
# ['Cat3', 'Cat1', 'Cat2']
expected_hierarchy = dict()
expected_hierarchy.update({None: self.cat3.id})
expected_hierarchy.update({self.cat3_uniq_le: self.cat1.id})
expected_hierarchy.update({self.cat3_uniq_gt: self.cat1.id})
expected_hierarchy.update({self.cat1_uniq_le: self.cat2.id})
expected_hierarchy.update({self.cat1_uniq_gt: self.cat2.id})
expected_hierarchy.update({self.cat2_uniq_gt: None})
expected_hierarchy.update({self.cat2_uniq_le: None})
self.assertEqual(expected_hierarchy, pkdss.hierarchy)
# [PKC1 PKC2]
# Cat3 [PKC1 PKC2]>0 []<=0
# Cat1 [PKC1]>2 [PKC2]<=2
# Cat2 [PKC1]>3 [PKC2]<=3
expected_symboltree = dict()
expected_symboltree.update(
{None: {
self.cat3_uniq_gt: ['PKC1', 'PKC2']
}})
expected_symboltree.update({
self.cat3_uniq_gt: {
self.cat1_uniq_gt: ['PKC1'],
self.cat1_uniq_le: ['PKC2']
}
})
expected_symboltree.update(
{self.cat1_uniq_gt: {
self.cat2_uniq_gt: ['PKC1']
}})
expected_symboltree.update(
{self.cat1_uniq_le: {
self.cat2_uniq_le: ['PKC2']
}})
self.assertEqual(expected_symboltree, pkdss.symboltree)
def test_init1(self):
sc = SubsetSizeCategorizer(self.size_thr)
self.assertItemsEqual(['>', '<='], sc.categories())