def test_get_ordered_idx(self):
mg = mini_gctoo_for_testing.make()
# case 1: id_type == None
case1 = parse_gctx.get_ordered_idx(None, [], mg.row_metadata_df)
self.assertEqual(
case1, range(0, 6),
"Expected oredered idx to be {} but got {}".format(
range(0, 6), case1))
# case 2: id_type == "id"
case2 = parse_gctx.get_ordered_idx(
"id", ['LJP007_MCF7_24H:CTL_VEHICLE:DMSO:-666'],
mg.col_metadata_df)
self.assertEqual(
case2, [4],
"Expected oredered idx to be {} but got {}".format([4], case2))
# case 3: id_type == ridx
case3 = parse_gctx.get_ordered_idx("idx", [5, 1, 3],
mg.col_metadata_df)
self.assertEqual(
case3, [1, 3, 5],
"Expected oredered idx to be {} but got {}".format([1, 3, 5],
case3))
def test_parse_metadata_df(self):
mini_gctoo = mini_gctoo_for_testing.make()
# convert row_metadata to np.nan
mini_row_meta = mini_gctoo.row_metadata_df.replace([-666, "-666", -666.0],
[np.nan, np.nan, np.nan])
gctx_file = h5py.File("functional_tests/mini_gctoo_for_testing.gctx", "r")
row_dset = gctx_file[row_meta_group_node]
col_dset = gctx_file[col_meta_group_node]
# with convert_neg_666
row_df = parse_gctx.parse_metadata_df("row", row_dset, True)
assert_frame_equal(mini_row_meta, row_df)
# no convert_neg_666
mini_gctoo_with_neg_666 = mini_gctoo_for_testing.make(convert_neg_666=False)
col_df = parse_gctx.parse_metadata_df("col", col_dset, False)
assert_frame_equal(mini_gctoo_with_neg_666.col_metadata_df, col_df)
def test_write_src(self):
# case 1: gctoo obj doesn't have src
mini1 = mini_gctoo_for_testing.make()
mini1.src = None
write_gctx.write(mini1, "no_src_example")
hdf5_file = h5py.File("no_src_example.gctx")
hdf5_src1 = hdf5_file.attrs[write_gctx.src_attr]
hdf5_file.close()
self.assertEqual(hdf5_src1, "no_src_example.gctx")
os.remove("no_src_example.gctx")
# case 2: gctoo obj does have src
mini2 = mini_gctoo_for_testing.make()
write_gctx.write(mini2, "with_src_example.gctx")
hdf5_file = h5py.File("with_src_example.gctx")
hdf5_src2 = hdf5_file.attrs[write_gctx.src_attr]
hdf5_file.close()
self.assertEqual(hdf5_src2, "mini_gctoo.gctx")
os.remove("with_src_example.gctx")
def test_write_version(self):
#TODO @oana refactor this test so it just calls the write_version method
# case 1: gctoo obj doesn't have version
mini1 = mini_gctoo_for_testing.make()
mini1.version = None
fn = "no_version_provided_example.gctx"
write_gctx.write(mini1, fn)
hdf5_file = h5py.File(fn)
hdf5_v1 = hdf5_file.attrs[write_gctx.version_attr]
hdf5_file.close()
self.assertEqual(hdf5_v1, write_gctx.version_number)
os.remove(fn)
# case 2: gctoo obj does have version, but it is not used when writing
mini2 = mini_gctoo_for_testing.make()
mini2.version = "MY_VERSION"
fn = "with_version_provided_example.gctx"
write_gctx.write(mini2, fn)
hdf5_file = h5py.File(fn)
hdf5_v2 = hdf5_file.attrs[write_gctx.version_attr]
hdf5_file.close()
self.assertEqual(hdf5_v2, write_gctx.version_number)
os.remove(fn)
def test_set_metadata_index_and_column_names(self):
mini_gctoo = mini_gctoo_for_testing.make()
mini_gctoo.row_metadata_df.index.name = None
mini_gctoo.row_metadata_df.columns.name = None
mini_gctoo.col_metadata_df.index.name = None
mini_gctoo.col_metadata_df.columns.name = None
# case 1: dim == "row"
parse_gctx.set_metadata_index_and_column_names("row", mini_gctoo.row_metadata_df)
self.assertEqual(mini_gctoo.row_metadata_df.index.name, "rid")
self.assertEqual(mini_gctoo.row_metadata_df.columns.name, "rhd")
# case 2: dim == "col"
parse_gctx.set_metadata_index_and_column_names("col", mini_gctoo.col_metadata_df)
self.assertEqual(mini_gctoo.col_metadata_df.index.name, "cid")
self.assertEqual(mini_gctoo.col_metadata_df.columns.name, "chd")
def test_make_specified_size_gctoo(self):
mini_gctoo = mini_gctoo_for_testing.make()
logger.debug("mini gctoo data_df shape: {}".format(mini_gctoo.data_df.shape))
logger.debug("mini gctoo row_meta shape: {}".format(mini_gctoo.row_metadata_df.shape))
logger.debug("mini gctoo col_meta shape: {}".format(mini_gctoo.col_metadata_df.shape))
# case 1: dim isn't 'row' or 'col'
with self.assertRaises(AssertionError) as context:
random_slice.make_specified_size_gctoo(mini_gctoo, 3, "aaaalll")
self.assertEqual(str(context.exception), "dim specified must be either 'row' or 'col'")
# case 2: row subsetting - happy
row_subset = random_slice.make_specified_size_gctoo(mini_gctoo, 3, "row")
self.assertEqual(row_subset.data_df.shape, (3,6),
"data_df after row slice is incorrect shape: {} vs (3,6)".format(row_subset.data_df.shape))
self.assertEqual(row_subset.row_metadata_df.shape, (3,5),
"row_metadata_df after row slice is incorrect shape: {} vs (3,5)".format(row_subset.row_metadata_df.shape))
self.assertEqual(row_subset.col_metadata_df.shape, (6,5),
"col_metadata_df after row slice is incorrect shape: {} vs (6,5)".format(row_subset.col_metadata_df.shape))
# case 3: row subsetting - sample subset > og # of samples
with self.assertRaises(AssertionError) as context:
random_slice.make_specified_size_gctoo(mini_gctoo, 30, "row")
self.assertEqual(str(context.exception), "number of samples must be subset of original file sample size")
# case 4: col subsetting - happy
col_subset = random_slice.make_specified_size_gctoo(mini_gctoo, 3, "col")
self.assertEqual(col_subset.data_df.shape, (6,3),
"data_df after col slice is incorrect shape: {} vs (6,3)".format(col_subset.data_df.shape))
self.assertEqual(col_subset.row_metadata_df.shape, (6, 5),
"row_metadata_df after col slice is incorrect shape: {} vs (6, 5)".format(col_subset.row_metadata_df.shape))
self.assertEqual(col_subset.col_metadata_df.shape, (3,5),
"col_metadata_df after col slice is incorrect shape: {} vs (3,5)".format(col_subset.col_metadata_df.shape))
# case 5: col subsetting - sample subset > og # of samples
with self.assertRaises(AssertionError) as context:
random_slice.make_specified_size_gctoo(mini_gctoo, 7, "col")
self.assertEqual(str(context.exception), "number of samples must be subset of original file sample size")
import h5py
import GCTXAttrInfo
import parse_gctoox
import write_gctoox
import mini_gctoo_for_testing
__author__ = "Oana Enache"
__email__ = "*****@*****.**"
FUNCTIONAL_TESTS_PATH = "functional_tests"
# instantiate logger
logger = logging.getLogger(setup_logger.LOGGER_NAME)
# instance of mini_gctoo for testing
mini_gctoo = mini_gctoo_for_testing.make()
version_node = "version"
rid_node = "/0/META/ROW/id"
cid_node = "/0/META/COL/id"
data_node = "/0/DATA/0/matrix"
row_meta_group_node = "/0/META/ROW"
col_meta_group_node = "/0/META/COL"
class TestParseGCTooX(unittest.TestCase):
def test_add_gctx_to_out_name(self):
name1 = "my_cool_file"
name2 = "my_other_cool_file.gctx"
# case 1: out file name doesn't end in gctx
def test_write_metadata(self):
"""
CASE 1:
- write metadata (has '-666') to file, do not convert -666
- parse in written metadata, don't convert -666
"""
mini_gctoo = mini_gctoo_for_testing.make(convert_neg_666=False)
hdf5_writer = h5py.File(
FUNCTIONAL_TESTS_PATH + "/mini_gctoo_metadata.gctx", "w")
write_gctx.write_metadata(hdf5_writer, "row",
mini_gctoo.row_metadata_df, False)
write_gctx.write_metadata(hdf5_writer, "col",
mini_gctoo.col_metadata_df, False)
hdf5_writer.close()
logger.debug("Wrote mini_gctoo_metadata.gctx to {}".format(
os.path.join(FUNCTIONAL_TESTS_PATH, "mini_gctoo_metadata.gctx")))
# read in written metadata, then close and delete file
mini_gctoo_col_metadata = parse_gctx.get_column_metadata(
FUNCTIONAL_TESTS_PATH + "/mini_gctoo_metadata.gctx",
convert_neg_666=False)
mini_gctoo_row_metadata = parse_gctx.get_row_metadata(
FUNCTIONAL_TESTS_PATH + "/mini_gctoo_metadata.gctx",
convert_neg_666=False)
os.remove(FUNCTIONAL_TESTS_PATH + "/mini_gctoo_metadata.gctx")
# check row metadata
self.assertTrue(
set(mini_gctoo.row_metadata_df.columns) == set(
mini_gctoo_row_metadata.columns),
"Mismatch between expected row metadata columns {} and column values written to file: {}"
.format(mini_gctoo.row_metadata_df.columns,
mini_gctoo_row_metadata.columns))
self.assertTrue(
set(mini_gctoo.row_metadata_df.index) == set(
mini_gctoo.col_metadata_df.index),
"Mismatch between expect row metadata index {} and index values written to file: {}"
.format(mini_gctoo.row_metadata_df.index,
mini_gctoo_row_metadata.index))
for c in list(mini_gctoo.row_metadata_df.columns):
logger.debug("C1: For column name: {}".format(c))
logger.debug("C1: populated values: {}".format(
set(mini_gctoo_row_metadata[c])))
logger.debug("C1: mini_gctoo values: {}".format(
set(mini_gctoo.row_metadata_df[c])))
self.assertTrue(
set(mini_gctoo.row_metadata_df[c]) == set(
mini_gctoo_row_metadata[c]),
"Values in column {} differ between expected metadata and written row metadata: {} vs {}"
.format(c, set(mini_gctoo.row_metadata_df[c]),
set(mini_gctoo_row_metadata[c])))
# check col metadata
self.assertTrue(
set(mini_gctoo.col_metadata_df.columns) == set(
mini_gctoo_col_metadata.columns),
"Mismatch between expected col metadata columns {} and column values written to file: {}"
.format(mini_gctoo.col_metadata_df.columns,
mini_gctoo_col_metadata.columns))
self.assertTrue(
set(mini_gctoo.col_metadata_df.index) == set(
mini_gctoo.col_metadata_df.index),
"Mismatch between expect col metadata index {} and index values written to file: {}"
.format(mini_gctoo.col_metadata_df.index,
mini_gctoo_col_metadata.index))
for c in list(mini_gctoo.col_metadata_df.columns):
self.assertTrue(
set(mini_gctoo.col_metadata_df[c]) == set(
mini_gctoo_col_metadata[c]),
"Values in column {} differ between expected metadata and written col metadata!"
.format(c))
"""
CASE 2:
- write metadata (has NaN, not '-666') to file, do convert NaN back to '-666'
- parse in written metadata, don't convert -666
"""
# first convert mini_gctoo's row & col metadata dfs -666s to NaN
converted_row_metadata = mini_gctoo.row_metadata_df.replace(
[-666, "-666", -666.0], [numpy.nan, numpy.nan, numpy.nan])
logger.debug("First row of converted_row_metadata: {}".format(
converted_row_metadata.iloc[0]))
converted_col_metadata = mini_gctoo.col_metadata_df.replace(
[-666, "-666", -666.0], [numpy.nan, numpy.nan, numpy.nan])
# write row and col metadata fields from mini_gctoo_for_testing instance to file
# Note this time does convert back to -666
hdf5_writer = h5py.File(
FUNCTIONAL_TESTS_PATH + "/mini_gctoo_metadata.gctx", "w")
write_gctx.write_metadata(hdf5_writer, "row", converted_row_metadata,
True)
write_gctx.write_metadata(hdf5_writer, "col", converted_col_metadata,
True)
hdf5_writer.close()
# read in written metadata, then close and delete file
mini_gctoo_col_metadata = parse_gctx.get_column_metadata(
FUNCTIONAL_TESTS_PATH + "/mini_gctoo_metadata.gctx",
convert_neg_666=False)
mini_gctoo_row_metadata = parse_gctx.get_row_metadata(
FUNCTIONAL_TESTS_PATH + "/mini_gctoo_metadata.gctx",
convert_neg_666=False)
os.remove(FUNCTIONAL_TESTS_PATH + "/mini_gctoo_metadata.gctx")
# check row metadata
self.assertTrue(
set(mini_gctoo.row_metadata_df.columns) == set(
mini_gctoo_row_metadata.columns),
"Mismatch between expected row metadata columns {} and column values written to file: {}"
.format(mini_gctoo.row_metadata_df.columns,
mini_gctoo_row_metadata.columns))
self.assertTrue(
set(mini_gctoo.row_metadata_df.index) == set(
mini_gctoo.col_metadata_df.index),
"Mismatch between expect row metadata index {} and index values written to file: {}"
.format(mini_gctoo.row_metadata_df.index,
mini_gctoo_row_metadata.index))
for c in list(mini_gctoo.row_metadata_df.columns):
logger.debug("C2: For column name: {}".format(c))
logger.debug("C2: populated values: {}".format(
set(mini_gctoo_row_metadata[c])))
logger.debug("C2: mini_gctoo values: {}".format(
set(mini_gctoo.row_metadata_df[c])))
self.assertTrue(
set(mini_gctoo.row_metadata_df[c]) == set(
mini_gctoo_row_metadata[c]),
"Values in column {} differ between expected metadata and written row metadata!"
.format(c))
# check col metadata
self.assertTrue(
set(mini_gctoo.col_metadata_df.columns) == set(
mini_gctoo_col_metadata.columns),
"Mismatch between expected col metadata columns {} and column values written to file: {}"
.format(mini_gctoo.col_metadata_df.columns,
mini_gctoo_col_metadata.columns))
self.assertTrue(
set(mini_gctoo.col_metadata_df.index) == set(
mini_gctoo.col_metadata_df.index),
"Mismatch between expect col metadata index {} and index values written to file: {}"
.format(mini_gctoo.col_metadata_df.index,
mini_gctoo_col_metadata.index))
for c in list(mini_gctoo.col_metadata_df.columns):
self.assertTrue(
set(mini_gctoo.col_metadata_df[c]) == set(
mini_gctoo_col_metadata[c]),
"Values in column {} differ between expected metadata and written col metadata!"
.format(c))
def test_parse(self):
# parse whole thing
mg1 = mini_gctoo_for_testing.make()
mg2 = parse_gctx.parse("functional_tests/mini_gctoo_for_testing.gctx")
assert_frame_equal(mg1.data_df, mg2.data_df)
assert_frame_equal(mg1.row_metadata_df, mg2.row_metadata_df)
assert_frame_equal(mg1.col_metadata_df, mg2.col_metadata_df)
# test with string rid/cid
test_rids = [
'LJP007_MCF10A_24H:TRT_CP:BRD-K93918653:3.33',
'LJP007_MCF7_24H:CTL_VEHICLE:DMSO:-666'
]
test_cids = ['LJP007_MCF7_24H:TRT_POSCON:BRD-A61304759:10']
mg3 = slice_gct.slice_gctoo(mg1, rid=test_rids, cid=test_cids)
mg4 = parse_gctx.parse("functional_tests/mini_gctoo_for_testing.gctx",
rid=test_rids,
cid=test_cids)
assert_frame_equal(mg3.data_df, mg4.data_df)
assert_frame_equal(mg3.row_metadata_df, mg4.row_metadata_df)
assert_frame_equal(mg3.col_metadata_df, mg4.col_metadata_df)
# first, make & write out temp version of mini_gctoo with int rids/cids
new_mg = mini_gctoo_for_testing.make(convert_neg_666=False)
int_indexed_data_df = new_mg.data_df.copy()
int_indexed_data_df.index = range(0, 6)
int_indexed_data_df.columns = range(10, 16)
int_indexed_row_meta = new_mg.row_metadata_df.copy()
int_indexed_row_meta.index = range(0, 6)
int_indexed_col_meta = new_mg.col_metadata_df.copy()
int_indexed_col_meta.index = range(10, 16)
int_indexed_gctoo = GCToo.GCToo(data_df=int_indexed_data_df,
row_metadata_df=int_indexed_row_meta,
col_metadata_df=int_indexed_col_meta)
write_gctx.write(int_indexed_gctoo, "int_indexed_mini_gctoo.gctx")
# test with numeric (repr as string) rid/cid
mg5 = GCToo.GCToo(data_df=int_indexed_data_df,
row_metadata_df=int_indexed_row_meta,
col_metadata_df=int_indexed_col_meta)
mg5 = slice_gct.slice_gctoo(
mg5,
row_bool=[True, False, True, False, True, False],
col_bool=[True, False, False, True, True, True])
mg5.data_df.index.name = "rid"
mg5.data_df.columns.name = "cid"
mg5.row_metadata_df.index.name = "rid"
mg5.row_metadata_df.columns.name = "rhd"
mg5.col_metadata_df.index.name = "cid"
mg5.col_metadata_df.columns.name = "chd"
mg6 = parse_gctx.parse("int_indexed_mini_gctoo.gctx",
rid=[0, 2, 4],
cid=[10, 13, 14, 15],
convert_neg_666=False)
os.remove("int_indexed_mini_gctoo.gctx")
assert_frame_equal(mg5.data_df, mg6.data_df)
assert_frame_equal(mg5.row_metadata_df, mg6.row_metadata_df)
assert_frame_equal(mg5.col_metadata_df, mg6.col_metadata_df)
# test with ridx/cidx
mg7 = slice_gct.slice_gctoo(
mg1,
rid=['LJP007_MCF7_24H:CTL_VEHICLE:DMSO:-666'],
cid='LJP007_MCF7_24H:CTL_VEHICLE:DMSO:-666')
mg8 = parse_gctx.parse("functional_tests/mini_gctoo_for_testing.gctx",
ridx=[4],
cidx=[4])
assert_frame_equal(mg7.data_df, mg8.data_df)
assert_frame_equal(mg7.row_metadata_df, mg8.row_metadata_df)
assert_frame_equal(mg7.col_metadata_df, mg8.col_metadata_df)
# test with rid/cidx
mg9 = parse_gctx.parse("functional_tests/mini_gctoo_for_testing.gctx",
rid=['LJP007_MCF7_24H:CTL_VEHICLE:DMSO:-666'],
cidx=[4])
assert_frame_equal(mg7.data_df, mg9.data_df)
assert_frame_equal(mg7.row_metadata_df, mg9.row_metadata_df)
assert_frame_equal(mg7.col_metadata_df, mg9.col_metadata_df)
# test with ridx/cid
mg10 = parse_gctx.parse("functional_tests/mini_gctoo_for_testing.gctx",
ridx=[4],
cid=['LJP007_MCF7_24H:CTL_VEHICLE:DMSO:-666'])
assert_frame_equal(mg7.data_df, mg10.data_df)
assert_frame_equal(mg7.row_metadata_df, mg10.row_metadata_df)
assert_frame_equal(mg7.col_metadata_df, mg10.col_metadata_df)
