def test_make_lmdb_dataset():
# Load PDB dataset
dataset = da.load_dataset('tests/test_data/pdb', 'pdb')
assert len(dataset) == 4
# Create LMDB dataset from PDB dataset
da.make_lmdb_dataset(dataset, 'tests/test_data/_output_lmdb',
filter_fn=None, serialization_format='json',
include_bonds=False)
# Try to load generated dataset
new_dataset = da.load_dataset('tests/test_data/_output_lmdb', 'lmdb')
assert len(new_dataset) == 4
# Remove temporary files
os.remove('tests/test_data/_output_lmdb/data.mdb')
os.remove('tests/test_data/_output_lmdb/lock.mdb')
os.rmdir('tests/test_data/_output_lmdb')
def test_load_dataset_xyz():
dataset = da.load_dataset('tests/test_data/xyz', 'xyz')
assert len(dataset) == 3
for df in dataset:
print(df)
assert df['atoms'].x.dtype == 'float'
assert df['atoms'].y.dtype == 'float'
assert df['atoms'].z.dtype == 'float'
def test_load_dataset_xyzgdb():
file_list = [
'tests/test_data/xyz-gdb/dsgdb9nsd_000005.xyz',
'tests/test_data/xyz-gdb/dsgdb9nsd_000212.xyz',
'tests/test_data/xyz-gdb/dsgdb9nsd_001458.xyz'
]
dataset = da.load_dataset(file_list, 'xyz-gdb')
assert len(dataset) == 3
def test_load_dataset_lmdb():
dataset = da.load_dataset('tests/test_data/lmdb', 'lmdb')
assert len(dataset) == 4
for df in dataset:
print(df)
assert df['atoms'].x.dtype == 'float'
assert df['atoms'].y.dtype == 'float'
assert df['atoms'].z.dtype == 'float'
def test_load_dataset_list():
dataset = da.load_dataset('tests/test_data/list/pdbs.txt', 'pdb')
assert len(dataset) == 4
for df in dataset:
print(df)
print(df['atoms'].x.dtype)
assert df['atoms'].x.dtype in ['float', 'float32', 'float64']
assert df['atoms'].y.dtype in ['float', 'float32', 'float64']
assert df['atoms'].z.dtype in ['float', 'float32', 'float64']
def test_load_dataset_sdf():
dataset = da.load_dataset('tests/test_data/sdf', 'sdf')
assert len(dataset) == 4
for df in dataset:
print(df)
print(df['atoms'].x.dtype)
assert df['atoms'].x.dtype in ['float', 'float32', 'float64']
assert df['atoms'].y.dtype in ['float', 'float32', 'float64']
assert df['atoms'].z.dtype in ['float', 'float32', 'float64']
def test_load_dataset_xyzgdb():
file_list = ['tests/test_data/xyz-gdb/dsgdb9nsd_000005.xyz',
'tests/test_data/xyz-gdb/dsgdb9nsd_000212.xyz',
'tests/test_data/xyz-gdb/dsgdb9nsd_001458.xyz']
dataset = da.load_dataset(file_list, 'xyz-gdb')
assert len(dataset) == 3
for df in dataset:
print(df)
assert df['atoms'].x.dtype == 'float'
assert df['atoms'].y.dtype == 'float'
assert df['atoms'].z.dtype == 'float'
def test_split():
# Load LMDB dataset
dataset = da.load_dataset('tests/test_data/lmdb', 'lmdb')
# Split with defined indices
indices_train, indices_val, indices_test = [3, 0], [2], [1]
s = spl.split(dataset, indices_train, indices_val, indices_test)
train_dataset, val_dataset, test_dataset = s
# Check whether the frames are in the correct dataset
assert dataset[0]['atoms'].equals(train_dataset[1]['atoms'])
assert dataset[1]['atoms'].equals(test_dataset[0]['atoms'])
assert dataset[2]['atoms'].equals(val_dataset[0]['atoms'])
assert dataset[3]['atoms'].equals(train_dataset[0]['atoms'])
def test_load_dataset_list_nonexistent():
dataset = da.load_dataset('tests/test_data/list/nonexistent.txt', 'pdb')
assert len(dataset) == 1
with pytest.raises(FileNotFoundError):
df = dataset[0]
def test_get_chain_sequences():
dataset = da.load_dataset('tests/test_data/lmdb', 'lmdb')
cseq = seq.get_chain_sequences(dataset[2]['atoms'])
assert cseq[0][1] == 'NNQQ'
def test_load_dataset_silent():
dataset = da.load_dataset('tests/test_data/silent', 'silent')
for x in dataset:
print(x)
assert len(dataset) == 8
import pandas as pd
import atom3d.datasets as da
import atom3d.filters.filters as filters
import atom3d.filters.scop as scop
dataset = da.load_dataset('tests/test_data/lmdb', 'lmdb')
# TODO: Commented code below fails.
# FileNotFoundError: [Errno 2] File ../../metadata/scop-cla-latest.txt does not exist: '../../metadata/scop-cla-latest.txt'
def test_scop_filter():
# level = 'class'
# filter_fn = scop.form_scop_filter(level, allowed=None, excluded=None)
# for i,d in enumerate(dataset):
# df_fil = filter_fn(d['atoms'])
pass
def load_example_dataset():
dataset = da.load_dataset(str(Path(__file__).parent.absolute()) + '/test_lmdb', 'lmdb')
return dataset
from atom3d.datasets import load_dataset, make_lmdb_dataset import sys in_path = sys.argv[1] out_path = sys.argv[2] dataset = load_dataset(in_path, 'lmdb') make_lmdb_dataset(dataset, out_path, filter_fn=lambda x: not x)
def test_load_dataset_sdf():
dataset = da.load_dataset('tests/test_data/sdf', 'sdf')
assert len(dataset) == 4
def test_load_dataset_silent():
dataset = da.load_dataset('tests/test_data/silent', 'silent')
assert len(dataset) == 4
def test_load_dataset_xyz():
dataset = da.load_dataset('tests/test_data/xyz-gdb', 'xyz')
assert len(dataset) == 3
def test_load_dataset_lmdb():
dataset = da.load_dataset('tests/test_data/lmdb', 'lmdb')
assert len(dataset) == 4
