def test_save_txt_errors(self):
"""Check errors/warnings raised when saving bitstrings."""
import warnings
from scipy.sparse import csr_matrix
from e3fp.util import E3FPEfficiencyWarning
from e3fp.fingerprint.db import FingerprintDatabase
from e3fp.fingerprint.fprint import CountFingerprint
from e3fp.fingerprint.util import E3FPInvalidFingerprintError
array = np.array(
[[1, 0, 0, 1, 1], [0, 0, 0, 1, 0], [0, 1, 1, 1, 1]], dtype=np.bool_
)
db = FingerprintDatabase.from_array(
array, ["1", "2", "3"], fp_type=CountFingerprint
)
desc, txt_file = tempfile.mkstemp(suffix=".txt.gz")
os.close(desc)
with self.assertRaises(E3FPInvalidFingerprintError):
db.savetxt(txt_file)
array = csr_matrix((3, 2 ** 15), dtype=np.bool_)
db = FingerprintDatabase.from_array(array, ["1", "2", "3"])
with warnings.catch_warnings(record=True):
warnings.simplefilter("error")
with self.assertRaises(E3FPEfficiencyWarning):
db.savetxt(txt_file)
os.unlink(txt_file)
def test_create_from_array(self):
from e3fp.fingerprint.fprint import Fingerprint, CountFingerprint
from e3fp.fingerprint.db import FingerprintDatabase
array = (np.random.uniform(0, 1, size=(10, 1024)) > 0.9).astype(
np.uint16
)
fprints = [Fingerprint.from_vector(array[i, :]) for i in range(10)]
fp_names = []
for i, fp in enumerate(fprints):
name = str(i)
fp.name = name
fp.level = 5
fp_names.append(name)
db1 = FingerprintDatabase(
fp_type=CountFingerprint, level=5, name="Test"
)
db1.add_fingerprints(fprints)
db2 = FingerprintDatabase.from_array(
array, fp_names, level=5, name="Test"
)
self.assertEqual(db1.fp_type, db2.fp_type)
np.testing.assert_array_equal(
db1.array.todense().getA(), db2.array.todense().getA()
)
def test_load_efficiency_warning(self):
import warnings
from e3fp.util import E3FPEfficiencyWarning
from e3fp.fingerprint.db import FingerprintDatabase
import scipy
array = (np.random.uniform(0, 1, size=(10, 1024)) > 0.9).astype(
np.uint16
)
fp_names = []
for i in range(array.shape[0]):
fp_names.append(str(i))
db = FingerprintDatabase.from_array(
array, fp_names=fp_names, level=5, props={"index": range(10)}
)
desc, db_file = tempfile.mkstemp(suffix=".fpz")
os.close(desc)
db.savez(db_file)
with warnings.catch_warnings(record=True):
warnings.simplefilter("error")
scipy.__version__ = "0.19"
with self.assertRaises(E3FPEfficiencyWarning):
db.load(db_file)
scipy.__version__ = "1.0"
db.load(db_file)
os.unlink(db_file)
def test_save_txt(self):
"""Ensure bitstrings saved to txt correctly."""
from e3fp.fingerprint.db import FingerprintDatabase
from python_utilities.io_tools import smart_open
array = np.array(
[[1, 0, 0, 1, 1], [0, 0, 0, 1, 0], [0, 1, 1, 1, 1]], dtype=np.bool_
)
db = FingerprintDatabase.from_array(array, ["1", "2", "3"])
desc, txt_file = tempfile.mkstemp(suffix=".txt.gz")
os.close(desc)
db.savetxt(txt_file)
exp_bitstring = b"10011 1\n00010 2\n01111 3\n"
with smart_open(txt_file, "r") as f:
bitstring = f.read()
self.assertEqual(bitstring, exp_bitstring)
os.unlink(txt_file)
desc, txt_file = tempfile.mkstemp(suffix=".txt.gz")
os.close(desc)
db.savetxt(txt_file, with_names=False)
exp_bitstring = b"10011\n00010\n01111\n"
with smart_open(txt_file, "r") as f:
bitstring = f.read()
self.assertEqual(bitstring, exp_bitstring)
os.unlink(txt_file)
def test_get_db_subset(self):
from e3fp.fingerprint.db import FingerprintDatabase
array = (np.random.uniform(0, 1, size=(10, 1024)) > .9).astype(
np.uint16)
fp_names = []
for i in range(array.shape[0]):
fp_names.append(str(i))
db = FingerprintDatabase.from_array(array, fp_names=fp_names, level=5)
sub_db = db.get_subset(fp_names[:-2])
self.assertEqual(db.level, sub_db.level)
self.assertEqual(db.fp_type, sub_db.fp_type)
self.assertEqual((db.array[:-2, :] - sub_db.array).nnz, 0)
def test_fingerprint_has_props(self):
from e3fp.fingerprint.db import FingerprintDatabase
array = (np.random.uniform(0, 1, size=(10, 1024)) > 0.9).astype(
np.uint16
)
fp_names = [str(i) for i in range(10)]
indices = [float(i) for i in range(10)]
db = FingerprintDatabase.from_array(
array, fp_names, level=5, name="Test", props={"index": indices}
)
for i, x in enumerate(fp_names):
self.assertEqual(db[x][0].get_prop("index"), indices[i])
def test_lookup(self):
from e3fp.fingerprint.fprint import Fingerprint
from e3fp.fingerprint.db import FingerprintDatabase
db = FingerprintDatabase(name="TestDB")
array = (np.random.uniform(0, 1, size=(2, 1024)) > .9).astype(np.bool_)
fp_names = []
for i in range(array.shape[0]):
fp_names.append("fp" + str(i))
db = FingerprintDatabase.from_array(array, fp_names, name="Test")
for i in range(array.shape[0]):
self.assertEqual(Fingerprint.from_vector(array[i, :]), db[i])
self.assertEqual(Fingerprint.from_vector(array[i, :]),
db[db.fp_names[i]][0])
def test_concat_dbs(self):
from e3fp.fingerprint.db import concat, FingerprintDatabase
array = (np.random.uniform(0, 1, size=(10, 1024)) > .9).astype(
np.uint16)
fp_names = [str(i) for i in range(10)]
indices = [float(i) for i in range(10)]
dbs = []
for i in range(10)[::2]:
db = FingerprintDatabase.from_array(
array[i:i + 2, :],
fp_names[i:i + 2],
level=5,
name="Test",
props={"index": indices[i:i + 2]})
dbs.append(db)
join_db = concat(dbs)
np.testing.assert_array_equal(join_db.array.todense().getA(), array)
def test_roundtrip_zlib(self):
"""Ensure DB is the same after saving with savez and loading."""
from e3fp.fingerprint.db import FingerprintDatabase
array = (np.random.uniform(0, 1, size=(10, 1024)) > .9).astype(
np.uint16)
fp_names = []
for i in range(array.shape[0]):
fp_names.append(str(i))
db = FingerprintDatabase.from_array(array,
fp_names=fp_names,
level=5,
props={"index": range(10)})
desc, db_file = tempfile.mkstemp(suffix=".fpz")
os.close(desc)
db.savez(db_file)
db2 = db.load(db_file)
os.unlink(db_file)
self.assertEqual(db, db2)
self.assertListEqual(db2.get_prop("index").tolist(), list(range(10)))