def test_main(self):
in_gct_path = os.path.join(FUNCTIONAL_TESTS_DIR, "test_slice_in.gct")
rid_grp_path = os.path.join(FUNCTIONAL_TESTS_DIR, "test_slice_rid.grp")
out_name = os.path.join(FUNCTIONAL_TESTS_DIR, "test_slice_out.gct")
expected_out_path = os.path.join(FUNCTIONAL_TESTS_DIR,
"test_slice_expected.gct")
args_string = "-i {} --rid {} -ec {} -o {}".format(
in_gct_path, rid_grp_path, "f", out_name)
args = slice_gct.build_parser().parse_args(args_string.split())
# Run main method
slice_gct.main(args)
# Compare output to expected
out_gct = pg.parse(out_name)
expected_gct = pg.parse(expected_out_path)
pd.util.testing.assert_frame_equal(out_gct.data_df,
expected_gct.data_df)
pd.util.testing.assert_frame_equal(out_gct.row_metadata_df,
expected_gct.row_metadata_df)
pd.util.testing.assert_frame_equal(out_gct.col_metadata_df,
expected_gct.col_metadata_df)
# Clean up
os.remove(out_name)
def test_concat_main(self):
test_dir = "../functional_tests/test_concat/test_main"
g_a = pg.parse(os.path.join(test_dir, "a.gct"))
logger.debug("g_a: {}".format(g_a))
g_b = pg.parse(os.path.join(test_dir, "b.gct"))
logger.debug("g_b: {}".format(g_b))
#unhappy path - write out error report file
expected_output_file = tempfile.mkstemp()[1]
logger.debug(
"unhappy path - write out error report file - expected_output_file: {}"
.format(expected_output_file))
args = cg.build_parser().parse_args([
"-d", "horiz", "-if", g_a.src, g_b.src, "-o", "should_not_be_used",
"-ot", "gct", "-erof", expected_output_file
])
logger.debug("args: {}".format(args))
with self.assertRaises(
cg.MismatchCommonMetadataConcatException) as context:
cg.concat_main(args)
self.assertTrue(os.path.exists(expected_output_file))
report_df = pd.read_csv(expected_output_file, sep="\t")
logger.debug("report_df:\n{}".format(report_df))
self.assertEqual(2, report_df.shape[0])
os.remove(expected_output_file)
#happy path
expected_output_file = tempfile.mkstemp(suffix=".gct")[1]
logger.debug("happy path - expected_output_file: {}".format(
expected_output_file))
args2 = cg.build_parser().parse_args([
"-d", "horiz", "-if", g_a.src, g_b.src, "-o", expected_output_file,
"-ot", "gct", "-ramf"
])
logger.debug("args2: {}".format(args2))
cg.concat_main(args2)
self.assertTrue(os.path.exists(expected_output_file))
r = pg.parse(expected_output_file)
logger.debug("happy path -r:\n{}".format(r))
logger.debug("r.data_df:\n{}".format(r.data_df))
self.assertEqual((2, 4), r.data_df.shape)
self.assertEqual({"a", "b", "g", "f"}, set(r.data_df.columns))
self.assertEqual({"rid1", "rid2"}, set(r.data_df.index))
#cleanup
os.remove(expected_output_file)
def test_top_bottom(self):
top_gct_path = os.path.join(FUNCTIONAL_TESTS_DIR, "test_merge_top.gct")
bottom_gct_path = os.path.join(FUNCTIONAL_TESTS_DIR, "test_merge_bottom.gct")
expected_gct_path = os.path.join(FUNCTIONAL_TESTS_DIR, "test_merged_top_bottom.gct")
top_gct = pg.parse(top_gct_path)
bottom_gct = pg.parse(bottom_gct_path)
expected_gct = pg.parse(expected_gct_path)
# Merge top and bottom
concated_gct = cg.vstack([top_gct, bottom_gct], [], False)
pd.util.testing.assert_frame_equal(expected_gct.data_df, concated_gct.data_df, check_names=False)
pd.util.testing.assert_frame_equal(expected_gct.row_metadata_df, concated_gct.row_metadata_df, check_names=False)
pd.util.testing.assert_frame_equal(expected_gct.col_metadata_df, concated_gct.col_metadata_df, check_names=False)
def test_left_right(self):
left_gct_path = os.path.join(FUNCTIONAL_TESTS_DIR, "test_merge_left.gct")
right_gct_path = os.path.join(FUNCTIONAL_TESTS_DIR, "test_merge_right.gct")
expected_gct_path = os.path.join(FUNCTIONAL_TESTS_DIR, "test_merged_left_right.gct")
left_gct = pg.parse(left_gct_path)
right_gct = pg.parse(right_gct_path)
expected_gct = pg.parse(expected_gct_path)
# Merge left and right
concated_gct = cg.hstack([left_gct, right_gct], [], False)
pd.util.testing.assert_frame_equal(expected_gct.data_df, concated_gct.data_df, check_names=False)
pd.util.testing.assert_frame_equal(expected_gct.row_metadata_df, concated_gct.row_metadata_df, check_names=False)
pd.util.testing.assert_frame_equal(expected_gct.col_metadata_df, concated_gct.col_metadata_df, check_names=False)
def test_main(self):
out_name = os.path.join(FUNCTIONAL_TESTS_PATH, "test_main_out.gct")
gctoo = GCToo.GCToo(data_df=self.data_df,
row_metadata_df=self.row_metadata_df,
col_metadata_df=self.col_metadata_df)
wg.write(gctoo,
out_name,
data_null="NaN",
metadata_null="-666",
filler_null="-666")
# Read in the gct and verify that it's the same as gctoo
new_gct = pg.parse(out_name)
pd.util.testing.assert_frame_equal(new_gct.data_df, gctoo.data_df)
pd.util.testing.assert_frame_equal(new_gct.row_metadata_df,
gctoo.row_metadata_df)
pd.util.testing.assert_frame_equal(new_gct.col_metadata_df,
gctoo.col_metadata_df)
# Also check that missing values were written to the file as expected
in_df = pd.read_csv(out_name,
sep="\t",
skiprows=2,
keep_default_na=False)
self.assertEqual(in_df.iloc[0, 1], "-666")
self.assertEqual(in_df.iloc[5, 6], "NaN")
# Cleanup
os.remove(out_name)
def test_parse_gct_int_ids(self):
path = os.path.join(FUNCTIONAL_TESTS_PATH, "test_parse_gct_int_ids.gct")
r = pg.parse(path)
logger.debug("r: {}".format(r))
logger.debug("r.data_df: {}".format(r.data_df))
self.assertEqual({"1", "2"}, set(r.data_df.columns))
self.assertEqual({"3", "11", "-3"}, set(r.data_df.index))
def test_p100_functional(self):
p100_in_path = os.path.join(FUNCTIONAL_TESTS_PATH, "test_p100.gct")
p100_out_path = os.path.join(FUNCTIONAL_TESTS_PATH, "test_p100_writing.gct")
# Read in original gct file
p100_in_gct = pg.parse(p100_in_path)
# Read in new gct file
wg.write(p100_in_gct, p100_out_path)
p100_out_gct = pg.parse(p100_out_path)
self.assertTrue(p100_in_gct.data_df.equals(p100_out_gct.data_df))
self.assertTrue(p100_in_gct.row_metadata_df.equals(p100_out_gct.row_metadata_df))
self.assertTrue(p100_in_gct.col_metadata_df.equals(p100_out_gct.col_metadata_df))
# Clean up
os.remove(p100_out_path)
def test_with_only_row_metadata(self):
# path to files
gctoo_path = FUNCTIONAL_TESTS_PATH + "/row_meta_only_example_n2x1203.gct"
gctoox_path = FUNCTIONAL_TESTS_PATH + "/row_meta_only_example_n2x1203.gctx"
# parse files
c2_gctoo = parse_gct.parse(gctoo_path)
c2_gctoox = parse_gctx.parse(gctoox_path)
# check rows and columns: data_df
self.assertTrue(set(list(c2_gctoo.data_df.index)) == set(list(c2_gctoox.data_df.index)),
"Mismatch between data_df index values of gct vs gctx: {} vs {}".format(c2_gctoo.data_df.index,
c2_gctoox.data_df.index))
self.assertTrue(set(list(c2_gctoo.data_df.columns)) == set(list(c2_gctoox.data_df.columns)),
"Mismatch between data_df column values of gct vs gctx: {} vs {}".format(
c2_gctoo.data_df.columns, c2_gctoox.data_df.columns))
logger.debug("c2 gctoo data_df columns equal to gctoox data_df columns? {}".format(
set(c2_gctoo.data_df.columns) == set(c2_gctoox.data_df.columns)))
for c in list(c2_gctoo.data_df.columns):
self.assertTrue(len(list(c2_gctoo.data_df[c])) == len(list(c2_gctoox.data_df[c])),
"Lengths of column {} differ between gct and gctx".format(c))
pandas_testing.assert_series_equal(c2_gctoo.data_df[c], c2_gctoox.data_df[c])
# check rows and columns: row_metadata_df
self.assertTrue(set(list(c2_gctoo.row_metadata_df.index)) == set(list(c2_gctoox.row_metadata_df.index)),
"Mismatch between row_metadata_df index values of gct vs gctx: {} vs {}".format(
c2_gctoo.row_metadata_df.index, c2_gctoox.row_metadata_df.index))
self.assertTrue(set(list(c2_gctoo.row_metadata_df.columns)) == set(list(c2_gctoox.row_metadata_df.columns)),
"Mismatch between row_metadata_df column values of gct vs gctx: {} vs {}".format(
c2_gctoo.row_metadata_df.columns, c2_gctoox.row_metadata_df.columns))
logger.debug("c2 gctoo row_metadata_df columns equal to gctoox row_metadata_df columns? {}".format(
set(c2_gctoo.row_metadata_df.columns) == set(c2_gctoox.row_metadata_df.columns)))
for c in list(c2_gctoo.row_metadata_df.columns):
self.assertTrue(len(list(c2_gctoo.row_metadata_df[c])) == len(list(c2_gctoox.row_metadata_df[c])),
"Lengths of column {} differ between gct and gctx".format(c))
self.assertTrue(c2_gctoo.row_metadata_df[c].dtype == c2_gctoox.row_metadata_df[c].dtype,
"Dtype mismatch between parsed gct & gctx: {} vs {}".format(
c2_gctoo.row_metadata_df[c].dtype, c2_gctoox.row_metadata_df[c].dtype))
logger.debug("first couple elems of {} in gctoo: {}".format(c, list(c2_gctoo.row_metadata_df[c])[0:3]))
pandas_testing.assert_series_equal(c2_gctoo.row_metadata_df[c], c2_gctoox.row_metadata_df[c])
# check rows and columns: col_metadata_df
self.assertTrue(set(list(c2_gctoo.col_metadata_df.index)) == set(list(c2_gctoox.col_metadata_df.index)),
"Mismatch between col_metadata_df index values of gct vs gctx: {} vs {}".format(
c2_gctoo.col_metadata_df.index, c2_gctoox.col_metadata_df.index))
self.assertTrue(set(list(c2_gctoo.col_metadata_df.columns)) == set(list(c2_gctoox.col_metadata_df.columns)),
"Mismatch between col_metadata_df column values of gct vs gctx: {} vs {}".format(
c2_gctoo.col_metadata_df.columns, c2_gctoox.col_metadata_df.columns))
logger.debug("c2 gctoo col_metadata_df columns equal to gctoox col_metadata_df columns? {}".format(
set(c2_gctoo.col_metadata_df.columns) == set(c2_gctoox.col_metadata_df.columns)))
for c in list(c2_gctoo.col_metadata_df.columns):
self.assertTrue(len(list(c2_gctoo.col_metadata_df[c])) == len(list(c2_gctoox.col_metadata_df[c])),
"Lengths of column {} differ between gct and gctx".format(c))
self.assertTrue(c2_gctoo.col_metadata_df[c].dtype == c2_gctoox.col_metadata_df[c].dtype,
"Dtype mismatch between parsed gct & gctx: {} vs {}".format(
c2_gctoo.col_metadata_df[c].dtype, c2_gctoox.col_metadata_df[c].dtype))
pandas_testing.assert_series_equal(c2_gctoo.col_metadata_df[c], c2_gctoox.col_metadata_df[c])
def test_gctx2gct_main(self):
in_name = "../functional_tests/mini_gctoo_for_testing.gctx"
out_name = "../functional_tests/test_gctx2gct_out.gct"
args_string = "-f {} -o {}".format(in_name, out_name)
args = gctx2gct.build_parser().parse_args(args_string.split())
gctx2gct.gctx2gct_main(args)
# Make sure the input is identical to output
in_gctx = parse_gctx.parse(in_name)
out_gct = parse_gct.parse(out_name)
pd.util.testing.assert_frame_equal(in_gctx.data_df,
out_gct.data_df,
check_less_precise=3)
pd.util.testing.assert_frame_equal(in_gctx.col_metadata_df,
out_gct.col_metadata_df)
pd.util.testing.assert_frame_equal(in_gctx.row_metadata_df,
out_gct.row_metadata_df)
no_meta = "../functional_tests/mini_gctoo_for_testing_nometa.gctx"
added_meta = "../functional_tests/test_gctx2gct_out_annotated.gct"
row_meta = "../functional_tests/test_rowmeta_n6.txt"
col_meta = "../functional_tests/test_colmeta_n6.txt"
args_string = "-f {} -o {} -row_annot_path {} -col_annot_path {}".format(
no_meta, added_meta, row_meta, col_meta)
args = gctx2gct.build_parser().parse_args(args_string.split())
gctx2gct.gctx2gct_main(args)
annotated_gct = parse_gct.parse(added_meta)
# Check added annotations are the same as original input GCTX
pd.util.testing.assert_frame_equal(in_gctx.data_df,
annotated_gct.data_df,
check_less_precise=3)
pd.util.testing.assert_frame_equal(in_gctx.col_metadata_df,
annotated_gct.col_metadata_df)
pd.util.testing.assert_frame_equal(in_gctx.row_metadata_df,
annotated_gct.row_metadata_df)
# Clean up
os.remove(out_name)
os.remove(added_meta)
def parse(file_path,
convert_neg_666=True,
rid=None,
cid=None,
ridx=None,
cidx=None,
row_meta_only=False,
col_meta_only=False,
make_multiindex=False):
"""
Identifies whether file_path corresponds to a .gct or .gctx file and calls the
correct corresponding parse method.
Input:
Mandatory:
- gct(x)_file_path (str): full path to gct(x) file you want to parse.
Optional:
- row_meta_only (bool): Whether to load data + metadata (if False), or just row metadata (if True)
as pandas DataFrame
- col_meta_only (bool): Whether to load data + metadata (if False), or just col metadata (if True)
as pandas DataFrame
- convert_neg_666 (bool): whether to convert -666 values to numpy.nan or not
(see Note below for more details on this). Default = False.
- rid (list of strings): list of row ids to specifically keep from gctx. Default=None.
- cid (list of strings): list of col ids to specifically keep from gctx. Default=None.
- make_multiindex (bool): whether to create a multi-index df combining
the 3 component dfs
Output:
- myGCToo (GCToo)
Note: why does convert_neg_666 exist?
- In CMap--for somewhat obscure historical reasons--we use "-666" as our null value
for metadata. However (so that users can take full advantage of pandas' methods,
including those for filtering nan's etc) we provide the option of converting these
into numpy.NaN values, the pandas default.
"""
if file_path.endswith(".gct"):
# Ignoring arguments that won't be passed to parse_gct
for unused_arg in ["rid", "cid", "ridx", "cidx"]:
if eval(unused_arg):
err_msg = "parse_gct does not use the argument {}. Ignoring it...".format(
unused_arg)
logger.error(err_msg)
raise Exception(err_msg)
curr = parse_gct.parse(file_path, convert_neg_666, row_meta_only,
col_meta_only, make_multiindex)
elif file_path.endswith(".gctx"):
curr = parse_gctx.parse(file_path, convert_neg_666, rid, cid, ridx,
cidx, row_meta_only, col_meta_only,
make_multiindex)
else:
err_msg = "File to parse must be .gct or .gctx!"
logger.error(err_msg)
raise Exception(err_msg)
return curr
def test_parse(self):
# L1000 gct
l1000_file_path = os.path.join(FUNCTIONAL_TESTS_PATH, "test_l1000.gct")
l1000_gct = pg.parse(l1000_file_path)
# Check a few values
self.assertAlmostEqual(l1000_gct.data_df.iloc[0, 0], 11.3819, places=4,
msg=("The first value in the data matrix should be " +
"{} not {}").format("11.3819", l1000_gct.data_df.iloc[0, 0]))
self.assertEqual(l1000_gct.col_metadata_df.iloc[0, 0], 58,
msg=("The first value in the column metadata should be " +
"{} not {}").format("58", l1000_gct.col_metadata_df.iloc[0, 0]))
self.assertEqual(l1000_gct.row_metadata_df.iloc[0, 0], "Analyte 11",
msg=("The first value in the row metadata should be " +
"{} not {}").format("Analyte 11", l1000_gct.row_metadata_df.iloc[0, 0]))
# P100 gct
p100_file_path = os.path.join(FUNCTIONAL_TESTS_PATH, "test_p100.gct")
p100_gct = pg.parse(p100_file_path)
# Check a few values
self.assertAlmostEqual(p100_gct.data_df.iloc[0, 0], 0.9182, places=4,
msg=("The first value in the data matrix should be " +
"{} not {}").format("0.9182", p100_gct.data_df.iloc[0, 0]))
self.assertEqual(p100_gct.col_metadata_df.iloc[0, 0], "MCF7",
msg=("The first value in the column metadata should be " +
"{} not {}").format("MCF7", p100_gct.col_metadata_df.iloc[0, 0]))
self.assertEqual(p100_gct.row_metadata_df.iloc[0, 0], 1859,
msg=("The first value in the row metadata should be " +
"{} not {}").format("1859", p100_gct.row_metadata_df.iloc[0, 0]))
# GCT1.2
gct_v1point2_path = os.path.join(FUNCTIONAL_TESTS_PATH, "test_v1point2_n5x10.gct")
gct_v1point2 = pg.parse(gct_v1point2_path)
# Check a few values
self.assertAlmostEqual(
gct_v1point2.data_df.loc["217140_s_at", "LJP005_A375_24H_X1_B19:A06"],
6.9966, places=4)
self.assertEqual(gct_v1point2.row_metadata_df.loc["203627_at", "Description"], "IGF1R")
# Make sure col_metadata_df is empty
self.assertEqual(gct_v1point2.col_metadata_df.size, 0,
"col_metadata_df should be empty.")
def parse(file_path, convert_neg_666=True, rid=None, cid=None, ridx=None, cidx=None,
row_meta_only=False, col_meta_only=False, make_multiindex=False):
"""
Identifies whether file_path corresponds to a .gct or .gctx file and calls the
correct corresponding parse method.
Input:
Mandatory:
- gct(x)_file_path (str): full path to gct(x) file you want to parse.
Optional:
- convert_neg_666 (bool): whether to convert -666 values to numpy.nan or not
(see Note below for more details on this). Default = False.
- rid (list of strings): list of row ids to specifically keep from gctx. Default=None.
- cid (list of strings): list of col ids to specifically keep from gctx. Default=None.
- ridx (list of integers): only read the rows corresponding to this
list of integer ids. Default=None.
- cidx (list of integers): only read the columns corresponding to this
list of integer ids. Default=None.
- row_meta_only (bool): Whether to load data + metadata (if False), or just row metadata (if True)
as pandas DataFrame
- col_meta_only (bool): Whether to load data + metadata (if False), or just col metadata (if True)
as pandas DataFrame
- make_multiindex (bool): whether to create a multi-index df combining
the 3 component dfs
Output:
- out (GCToo object or pandas df): if row_meta_only or col_meta_only, then
out is a metadata df; otherwise, it's a GCToo instance containing
content of parsed gct(x) file
Note: why does convert_neg_666 exist?
- In CMap--for somewhat obscure historical reasons--we use "-666" as our null value
for metadata. However (so that users can take full advantage of pandas' methods,
including those for filtering nan's etc) we provide the option of converting these
into numpy.NaN values, the pandas default.
"""
if file_path.endswith(".gct"):
out = parse_gct.parse(file_path, convert_neg_666=convert_neg_666,
rid=rid, cid=cid, ridx=ridx, cidx=cidx,
row_meta_only=row_meta_only, col_meta_only=col_meta_only,
make_multiindex=make_multiindex)
elif file_path.endswith(".gctx"):
out = parse_gctx.parse(file_path, convert_neg_666=convert_neg_666,
rid=rid, cid=cid, ridx=ridx, cidx=cidx,
row_meta_only=row_meta_only, col_meta_only=col_meta_only,
make_multiindex=make_multiindex)
else:
err_msg = "File to parse must be .gct or .gctx!"
logger.error(err_msg)
raise Exception(err_msg)
return out
def gct2gctx_main(args):
""" Separate from main() in order to make command-line tool. """
in_gctoo = parse_gct.parse(args.filename, convert_neg_666=False)
if args.output_filepath is None:
basename = os.path.basename(args.filename)
out_name = os.path.splitext(basename)[0] + ".gctx"
else:
out_name = args.output_filepath
write_gctx.write(in_gctoo, out_name)
def main():
args = build_parser().parse_args(sys.argv[1:])
setup_logger.setup(verbose=args.verbose)
in_gctoo = parse_gct.parse(args.filename, convert_neg_666=False)
logger.debug("Original out name: {}".format(in_gctoo.src))
if args.output_filepath == None:
out_name = str.split(in_gctoo.src, "/")[-1].split(".")[0]
else:
out_name = args.output_filepath
write_gctx.write(in_gctoo, out_name)
def test_l1000_functional(self):
l1000_in_path = os.path.join(FUNCTIONAL_TESTS_PATH, "test_l1000.gct")
l1000_out_path = os.path.join(FUNCTIONAL_TESTS_PATH,
"test_l1000_writing.gct")
# Read in original gct file
l1000_in_gct = pg.parse(l1000_in_path)
# do write operation
wg.write(l1000_in_gct, l1000_out_path)
# Read in new gct file
l1000_out_gct = pg.parse(l1000_out_path)
pd.testing.assert_frame_equal(l1000_in_gct.data_df,
l1000_out_gct.data_df)
pd.testing.assert_frame_equal(l1000_in_gct.row_metadata_df,
l1000_out_gct.row_metadata_df)
pd.testing.assert_frame_equal(l1000_in_gct.col_metadata_df,
l1000_out_gct.col_metadata_df)
# Clean up
os.remove(l1000_out_path)
def test_p100_functional(self):
p100_in_path = os.path.join(FUNCTIONAL_TESTS_PATH, "test_p100.gct")
p100_out_path = os.path.join(FUNCTIONAL_TESTS_PATH,
"test_p100_writing.gct")
# Read in original gct file
p100_in_gct = pg.parse(p100_in_path)
# do write operation - note data_float_format set to None to preserve precision of input file
wg.write(p100_in_gct, p100_out_path, data_float_format=None)
# Read in new gct file
p100_out_gct = pg.parse(p100_out_path)
pd.testing.assert_frame_equal(p100_in_gct.data_df,
p100_out_gct.data_df)
pd.testing.assert_frame_equal(p100_in_gct.row_metadata_df,
p100_out_gct.row_metadata_df)
pd.testing.assert_frame_equal(p100_in_gct.col_metadata_df,
p100_out_gct.col_metadata_df)
# Clean up
os.remove(p100_out_path)
def test_gct2gctx_main(self):
in_name = "functional_tests/mini_gctoo_for_testing.gct"
out_name = "functional_tests/test_gct2gctx_out.gctx"
args_string = "-f {} -o {}".format(in_name, out_name)
args = gct2gctx.build_parser().parse_args(args_string.split())
gct2gctx.gct2gctx_main(args)
# Make sure the input is identical to output
in_gct = parse_gct.parse(in_name)
out_gctx = parse_gctx.parse(out_name)
pd.util.testing.assert_frame_equal(in_gct.data_df, out_gctx.data_df)
pd.util.testing.assert_frame_equal(in_gct.col_metadata_df,
out_gctx.col_metadata_df)
pd.util.testing.assert_frame_equal(in_gct.row_metadata_df,
out_gctx.row_metadata_df)
# Clean up
os.remove(out_name)
def main():
# get args
args = build_parser().parse_args(sys.argv[1:])
setup_logger.setup(verbose=args.verbose)
# Read the input gct
in_gct = pg.parse(args.in_gct_path)
# Read in each of the command line arguments
rid = _read_arg(args.rid)
cid = _read_arg(args.cid)
exclude_rid = _read_arg(args.exclude_rid)
exclude_cid = _read_arg(args.exclude_cid)
# Slice the gct
out_gct = slice_gctoo(in_gct, rid=rid, cid=cid, exclude_rid=exclude_rid, exclude_cid=exclude_cid)
assert out_gct.data_df.size > 0, "Slicing yielded an empty gct!"
# Write the output gct
wg.write(out_gct, args.out_name, data_null="NaN", metadata_null="NA", filler_null="NA")
def subset_main(args):
""" Separate method from main() in order to make testing easier and to
enable command-line access. """
# Read in each of the command line arguments
rid = _read_arg(args.rid)
cid = _read_arg(args.cid)
exclude_rid = _read_arg(args.exclude_rid)
exclude_cid = _read_arg(args.exclude_cid)
# If GCT, use subset_gctoo
if args.in_path.endswith(".gct"):
in_gct = parse_gct.parse(args.in_path)
out_gct = sg.subset_gctoo(in_gct,
rid=rid,
cid=cid,
exclude_rid=exclude_rid,
exclude_cid=exclude_cid)
# If GCTx, use parse_gctx
else:
if (exclude_rid is not None) or (exclude_cid is not None):
msg = "exclude_{rid,cid} args not currently supported for parse_gctx."
raise (Exception(msg))
logger.info("Using hyperslab selection functionality of parse_gctx...")
out_gct = parse_gctx.parse(args.in_path, rid=rid, cid=cid)
# Write the output gct
if args.out_type == "gctx":
wgx.write(out_gct, args.out_name)
else:
wg.write(out_gct,
args.out_name,
data_null="NaN",
metadata_null="NA",
filler_null="NA")
def test_parse_gct(self):
# tests the parsing of a gct file with high precision values
gct_filepath = os.path.join(FUNCTIONAL_TESTS_PATH, 'test_l1000_highprecision.gct')
data_gct = parse_gct.parse(gct_filepath)
(data, row_metadata, col_metadata) = (data_gct.data_df, data_gct.row_metadata_df, data_gct.col_metadata_df)
e_dims = [978, 377, 11, 35]
actual_version = 'GCT1.3'
# Check shapes of outputs
self.assertTrue(row_metadata.shape == (e_dims[0], e_dims[2]),
("row_metadata.shape = {} " +
"but expected it to be ({}, {})").format(row_metadata.shape,
e_dims[0], e_dims[2]))
self.assertTrue(col_metadata.shape == (e_dims[1], e_dims[3]),
("col_metadata.shape = {} " +
"but expected it to be ({}, {})").format(col_metadata.shape,
e_dims[1], e_dims[3]))
self.assertTrue(data.shape == (e_dims[0], e_dims[1]),
("data.shape = {} " +
"but expected it to be ({}, {})").format(data.shape,
e_dims[0], e_dims[1]))
# Check version
self.assertEqual(actual_version, data_gct.version)
# Check the type of data
self.assertTrue(isinstance(data.iloc[0, 0], np.float32),
"The data should be a float32, not {}".format(type(data.iloc[0, 0])))
# Check a few high precision floating values in data
correct_val = np.float32(11.574655)
self.assertTrue(data.iloc[0, 0] == correct_val,
("The first value in the data matrix should be " +
"{} not {}").format(correct_val, data.iloc[0, 0]))
correct_val = np.float32(5.3183546)
self.assertTrue(data.iloc[e_dims[0] - 1, e_dims[1] - 1] == correct_val,
("The last value in the data matrix should be " +
str(correct_val) + " not {}").format(data.iloc[e_dims[0] - 1, e_dims[1] - 1]))
def gct2gctx_main(args):
""" Separate from main() in order to make command-line tool. """
in_gctoo = parse_gct.parse(args.filename, convert_neg_666=False)
if args.output_filepath is None:
basename = os.path.basename(args.filename)
out_name = os.path.splitext(basename)[0] + ".gctx"
else:
out_name = args.output_filepath
""" If annotations are supplied, parse table and set metadata_df """
if args.row_annot_path is None:
pass
else:
row_metadata = pd.read_csv(args.row_annot_path,
sep='\t',
index_col=0,
header=0,
low_memory=False)
assert all(in_gctoo.data_df.index.isin(row_metadata.index)), \
"Row ids in matrix missing from annotations file"
in_gctoo.row_metadata_df = row_metadata.loc[row_metadata.index.isin(
in_gctoo.data_df.index)]
if args.col_annot_path is None:
pass
else:
col_metadata = pd.read_csv(args.col_annot_path,
sep='\t',
index_col=0,
header=0,
low_memory=False)
assert all(in_gctoo.data_df.columns.isin(col_metadata.index)), \
"Column ids in matrix missing from annotations file"
in_gctoo.col_metadata_df = col_metadata.loc[col_metadata.index.isin(
in_gctoo.data_df.columns)]
write_gctx.write(in_gctoo, out_name)
def test_main(self):
test_dir = "functional_tests/test_concat_gctoo/test_main"
g_a = pg.parse(os.path.join(test_dir, "a.gct"))
logger.debug("g_a: {}".format(g_a))
g_b = pg.parse(os.path.join(test_dir, "b.gct"))
logger.debug("g_b: {}".format(g_b))
save_build_parser = cg.build_parser
class MockParser:
def __init__(self, args):
self.args = args
def parse_args(self, unused):
return self.args
#unhappy path - write out error report file
expected_output_file = tempfile.mkstemp()[1]
logger.debug(
"unhappy path - write out error report file - expected_output_file: {}"
.format(expected_output_file))
args = save_build_parser().parse_args([
"-d", "horiz", "-if", g_a.src, g_b.src, "-o", "should_not_be_used",
"-ot", "gct", "-erof", expected_output_file
])
logger.debug("args: {}".format(args))
my_mock_parser = MockParser(args)
cg.build_parser = lambda: my_mock_parser
with self.assertRaises(
cg.MismatchCommonMetadataConcatGctooException) as context:
cg.main()
self.assertTrue(os.path.exists(expected_output_file))
report_df = pd.read_csv(expected_output_file, sep="\t")
logger.debug("report_df:\n{}".format(report_df))
self.assertEqual(2, report_df.shape[0])
os.remove(expected_output_file)
print()
print()
print()
#happy path
args.remove_all_metadata_fields = True
args.error_report_output_file = None
expected_output_file = tempfile.mkstemp(suffix=".gct")[1]
logger.debug("happy path - expected_output_file: {}".format(
expected_output_file))
args.out_name = expected_output_file
my_mock_parser = MockParser(args)
cg.buid_parser = lambda: my_mock_parser
cg.main()
self.assertTrue(os.path.exists(expected_output_file))
r = pg.parse(expected_output_file)
logger.debug("happy path -r:\n{}".format(r))
logger.debug("r.data_df:\n{}".format(r.data_df))
self.assertEqual((2, 4), r.data_df.shape)
self.assertEqual({"a", "b", "g", "f"}, set(r.data_df.columns))
self.assertEqual({"rid1", "rid2"}, set(r.data_df.index))
#cleanup
os.remove(expected_output_file)
cg.build_parser = save_build_parser
def test_parse(self):
# L1000 gct
l1000_file_path = os.path.join(FUNCTIONAL_TESTS_PATH, "test_l1000.gct")
l1000_gct = pg.parse(l1000_file_path)
logger.debug("parse L1000 gct - l1000_gct: {}".format(l1000_gct))
self.assertEqual(GCToo.GCToo, type(l1000_gct))
# Check a few values
self.assertAlmostEqual(l1000_gct.data_df.iloc[0, 0], 11.3819, places=4,
msg=("The first value in the data matrix should be " +
"{} not {}").format("11.3819", l1000_gct.data_df.iloc[0, 0]))
self.assertEqual(l1000_gct.col_metadata_df.iloc[0, 0], 58,
msg=("The first value in the column metadata should be " +
"{} not {}").format("58", l1000_gct.col_metadata_df.iloc[0, 0]))
self.assertEqual(l1000_gct.row_metadata_df.iloc[0, 0], "Analyte 11",
msg=("The first value in the row metadata should be " +
"{} not {}").format("Analyte 11", l1000_gct.row_metadata_df.iloc[0, 0]))
# P100 gct
p100_file_path = os.path.join(FUNCTIONAL_TESTS_PATH, "test_p100.gct")
p100_gct = pg.parse(p100_file_path)
logger.debug("parse P100 gct - p100_gct: {}".format(p100_gct))
self.assertEqual(GCToo.GCToo, type(p100_gct))
# Check a few values
self.assertAlmostEqual(p100_gct.data_df.iloc[0, 0], 0.9182, places=4,
msg=("The first value in the data matrix should be " +
"{} not {}").format("0.9182", p100_gct.data_df.iloc[0, 0]))
self.assertEqual(p100_gct.col_metadata_df.iloc[0, 0], "MCF7",
msg=("The first value in the column metadata should be " +
"{} not {}").format("MCF7", p100_gct.col_metadata_df.iloc[0, 0]))
self.assertEqual(p100_gct.row_metadata_df.iloc[0, 0], 1859,
msg=("The first value in the row metadata should be " +
"{} not {}").format("1859", p100_gct.row_metadata_df.iloc[0, 0]))
# GCT1.2
gct_v1point2_path = os.path.join(FUNCTIONAL_TESTS_PATH, "test_v1point2_n5x10.gct")
gct_v1point2 = pg.parse(gct_v1point2_path)
logger.debug("parse GC1.2 - gct_v1point2: {}".format(gct_v1point2))
self.assertEqual(GCToo.GCToo, type(gct_v1point2))
# Check a few values
self.assertAlmostEqual(
gct_v1point2.data_df.loc["217140_s_at", "LJP005_A375_24H_X1_B19:A06"],
6.9966, places=4)
self.assertEqual(gct_v1point2.row_metadata_df.loc["203627_at", "Description"], "IGF1R")
# Make sure col_metadata_df is empty
self.assertEqual(gct_v1point2.col_metadata_df.size, 0,
"col_metadata_df should be empty.")
#row_meta_only=True
row_metadata_df = pg.parse(l1000_file_path, row_meta_only=True)
logger.debug("row_meta_only=True - row_metadata_df.shape: {}".format(row_metadata_df.shape))
self.assertEqual(pd.DataFrame, type(row_metadata_df))
self.assertEqual((978, 11), row_metadata_df.shape)
#col_meta_only=True
col_metadata_df = pg.parse(l1000_file_path, col_meta_only=True)
logger.debug("col_meta_only=True - col_metadata_df.shape: {}".format(col_metadata_df.shape))
self.assertEqual(pd.DataFrame, type(col_metadata_df))
self.assertEqual((377, 35), col_metadata_df.shape)
def eVIP_run_main(pred_file=None,
sig_info=None,
gctx=None,
sig_gctx=None,
ref_allele_mode=None,
null_conn=None,
out_dir=None,
ymin=None,
ymax=None,
allele_col=None,
use_c_pval=None,
pdf=None,
cell_id=None,
plate_id=None,
corr_val_str=None):
#setting default values
ymin = int(ymin) if ymin != None else int(-1.00)
ymax = int(ymax) if ymax != None else int(1.00)
pred_file = open(pred_file)
pred_col = "prediction"
if os.path.exists(out_dir):
out_dir = os.path.abspath(out_dir)
else:
os.mkdir(out_dir)
out_dir = os.path.abspath(out_dir)
sig_info = open(sig_info)
null_conn = getNullConnDist(null_conn)
this_gctx = gct.parse(gctx)
sig_gctx = gct.parse(sig_gctx)
(gene2wt, gene2allele_call, gene2num_alleles,
allele2pvals) = parse_pred_file(pred_file, pred_col, use_c_pval,
ref_allele_mode)
allele2distil_ids = parse_sig_info(sig_info, allele_col, cell_id, plate_id)
for gene in gene2wt:
this_fig = plt.figure()
this_fig.set_size_inches((gene2num_alleles[gene] + 1) * 4, 4 * 3)
grid_size = (4, gene2num_alleles[gene] + 1)
wt_heatmap_ax = plt.subplot2grid(grid_size, (0, 0))
wt_im = plot_rep_heatmap(wt_heatmap_ax, this_gctx.data_df,
allele2distil_ids[gene2wt[gene]],
allele2distil_ids[gene2wt[gene]],
gene2wt[gene], ymin, ymax)
# WT self connectivity
wt_self, wt_self_row_medians = getSelfConnectivity(
this_gctx, allele2distil_ids[gene2wt[gene]],
len(allele2distil_ids[gene2wt[gene]]))
# Create consistent x values for the wt reps when plotting
wt_x_vals = []
for val in wt_self_row_medians:
wt_x_vals.append(random.randint(WT_RANGE[0], WT_RANGE[1]))
# Plot color bar on this axis
plt.colorbar(wt_im, ax=wt_heatmap_ax, shrink=0.7)
# Plot allele data
col_counter = 1
for type in PRED_TYPE:
for allele in gene2allele_call[gene][type]:
# CREATE SCATTERPLOT FIGURE
plot_signatures(pdf, out_dir, sig_gctx.data_df, gene2wt[gene],
allele, allele2distil_ids[gene2wt[gene]],
allele2distil_ids[allele])
# PLOT HEATMAP
this_hm_ax = plt.subplot2grid(grid_size, (0, col_counter))
plot_rep_heatmap(this_hm_ax, this_gctx.data_df,
allele2distil_ids[allele],
allele2distil_ids[allele],
type + " - " + allele, ymin, ymax)
# PLOT WT MUT heatmap
this_wt_mut_ax = plt.subplot2grid(grid_size, (1, col_counter))
plot_rep_heatmap(this_wt_mut_ax, this_gctx.data_df,
allele2distil_ids[gene2wt[gene]],
allele2distil_ids[allele],
gene2wt[gene] + " vs " + allele, ymin, ymax)
# PLOT RANKPOINT ROWS
this_jitter_ax = plt.subplot2grid(grid_size, (2, col_counter))
mut_self, mt_self_row_medians = getSelfConnectivity(
this_gctx, allele2distil_ids[allele],
len(allele2distil_ids[allele]))
wt_mut, wt_mut_row_medians = getConnectivity(
this_gctx, allele2distil_ids[gene2wt[gene]],
allele2distil_ids[allele], len(allele2distil_ids[allele]))
plot_jitter(
this_jitter_ax,
col_counter,
wt_x_vals,
wt_self_row_medians,
mt_self_row_medians,
wt_mut_row_medians,
# null_x_vals,
# null_conn,
allele2pvals[allele][0],
allele2pvals[allele][1],
use_c_pval,
ymin,
ymax,
corr_val_str)
# Compared to random connectivity
conn_ax = plt.subplot2grid(grid_size, (3, col_counter))
plot_conn(conn_ax, col_counter, null_conn, wt_mut_row_medians,
allele2pvals[allele][2], use_c_pval, corr_val_str)
col_counter += 1
if pdf:
this_fig.savefig("%s/%s_impact_pred_plots.pdf" % (out_dir, gene),
format="pdf")
else:
this_fig.savefig("%s/%s_impact_pred_plots.png" % (out_dir, gene))
plt.close(this_fig)
def test_with_both_metadata_fields(self):
# path to files
gctoo_path = FUNCTIONAL_TESTS_PATH + "/both_metadata_example_n1476x978.gct"
gctoox_path = FUNCTIONAL_TESTS_PATH + "/both_metadata_example_n1476x978.gctx"
# parse files
c1_gctoo = parse_gct.parse(gctoo_path)
c1_gctoox = parse_gctx.parse(gctoox_path)
# check rows and columns: data_df
self.assertTrue(set(list(c1_gctoo.data_df.index)) == set(list(c1_gctoox.data_df.index)),
"Mismatch between data_df index values of gct vs gctx: {} vs {}".format(c1_gctoo.data_df.index,
c1_gctoox.data_df.index))
self.assertTrue(set(list(c1_gctoo.data_df.columns)) == set(list(c1_gctoox.data_df.columns)),
"Mismatch between data_df column values of gct vs gctx: {} vs {}".format(
c1_gctoo.data_df.columns, c1_gctoox.data_df.columns))
logger.debug("c1 gctoo data_df columns equal to gctoox data_df columns? {}".format(
set(c1_gctoo.data_df.columns) == set(c1_gctoox.data_df.columns)))
for c in list(c1_gctoo.data_df.columns):
# logger.debug("Comparing data values in Column: {}".format(c))
self.assertTrue(len(list(c1_gctoo.data_df[c])) == len(list(c1_gctoox.data_df[c])),
"Lengths of column {} differ between gct and gctx".format(c))
pandas_testing.assert_series_equal(c1_gctoo.data_df[c], c1_gctoox.data_df[c])
# check rows and columns: row_metadata_df
self.assertTrue(set(list(c1_gctoo.row_metadata_df.index)) == set(list(c1_gctoox.row_metadata_df.index)),
"Mismatch between row_metadata_df index values of gct vs gctx: {} vs {}".format(
c1_gctoo.row_metadata_df.index, c1_gctoox.row_metadata_df.index))
self.assertTrue(set(list(c1_gctoo.row_metadata_df.columns)) == set(list(c1_gctoox.row_metadata_df.columns)),
"Mismatch between row_metadata_df column values of gct vs gctx: difference is {}".format(
set(c1_gctoo.row_metadata_df.columns).symmetric_difference(
set(c1_gctoox.row_metadata_df.columns))))
logger.debug("c1 gctoo row_metadata_df columns equal to gctoox row_metadata_df columns? {}".format(
set(c1_gctoo.row_metadata_df.columns) == set(c1_gctoox.row_metadata_df.columns)))
logger.debug("c1 gctoo dtypes: {}".format(c1_gctoo.row_metadata_df.dtypes))
logger.debug("c1 gctoox dtypes: {}".format(c1_gctoox.row_metadata_df.dtypes))
for c in list(c1_gctoo.row_metadata_df.columns):
self.assertTrue(len(list(c1_gctoo.row_metadata_df[c])) == len(list(c1_gctoox.row_metadata_df[c])),
"Lengths of column {} differ between gct and gctx".format(c))
logger.debug("first couple elems of {} in gctoo: {}".format(c, list(c1_gctoo.row_metadata_df[c])[0:3]))
self.assertTrue(c1_gctoo.row_metadata_df[c].dtype == c1_gctoox.row_metadata_df[c].dtype,
"Dtype mismatch for {} between parsed gct & gctx: {} vs {}".format(c,
c1_gctoo.row_metadata_df[
c].dtype,
c1_gctoox.row_metadata_df[
c].dtype))
pandas_testing.assert_series_equal(c1_gctoo.row_metadata_df[c], c1_gctoox.row_metadata_df[c])
# check rows and columns: col_metadata_df
self.assertTrue(set(list(c1_gctoo.col_metadata_df.index)) == set(list(c1_gctoox.col_metadata_df.index)),
"Mismatch between col_metadata_df index values of gct vs gctx: {} vs {}".format(
c1_gctoo.col_metadata_df.index, c1_gctoox.col_metadata_df.index))
self.assertTrue(set(list(c1_gctoo.col_metadata_df.columns)) == set(list(c1_gctoox.col_metadata_df.columns)),
"Mismatch between col_metadata_df column values of gct vs gctx: {} vs {}".format(
c1_gctoo.col_metadata_df.columns, c1_gctoox.col_metadata_df.columns))
logger.debug("c1 gctoo col_metadata_df columns equal to gctoox col_metadata_df columns? {}".format(
set(c1_gctoo.col_metadata_df.columns) == set(c1_gctoox.col_metadata_df.columns)))
for c in list(c1_gctoo.col_metadata_df.columns):
self.assertTrue(len(list(c1_gctoo.col_metadata_df[c])) == len(list(c1_gctoox.col_metadata_df[c])),
"Lengths of column {} differ between gct and gctx".format(c))
self.assertTrue(c1_gctoo.col_metadata_df[c].dtype == c1_gctoox.col_metadata_df[c].dtype,
"Dtype mismatch between parsed gct & gctx: {} vs {}".format(
c1_gctoo.col_metadata_df[c].dtype, c1_gctoox.col_metadata_df[c].dtype))
pandas_testing.assert_series_equal(c1_gctoo.col_metadata_df[c], c1_gctoox.col_metadata_df[c])
mol_names = []
for filename in gct_files:
mol_names.append(filename.split(".gct")[0])
def process(data):
df = data.data_df.filter(like='24H', axis=1)
if not len(df.columns) == 1:
df = df.filter(like='10', axis=1)
#df = df.filter(like='CPC0', axis=1)
else:
pass
return df
col_names = []
dfs = []
for i in range(0, len(gct_files)):
data = parse(gct_files[i])
df = process(data)
df = pd.DataFrame(df.mean(axis=1))
if len(df.columns) > 0:
col_names.append(mol_names[i])
else:
pass
dfs.append(df)
dfs = [df.set_index(data.row_metadata_df['pr_gene_symbol']) for df in dfs]
df = pd.concat(dfs, axis=1)
df.columns = col_names
df = df.head(978)
def test_parse(self):
# L1000 gct
l1000_file_path = os.path.join(FUNCTIONAL_TESTS_PATH, "test_l1000.gct")
l1000_gct = pg.parse(l1000_file_path)
logger.debug("parse L1000 gct - l1000_gct: {}".format(l1000_gct))
self.assertEqual(GCToo.GCToo, type(l1000_gct))
# Check a few values
self.assertAlmostEqual(l1000_gct.data_df.iloc[0, 0], 11.3819, places=4,
msg=("The first value in the data matrix should be " +
"{} not {}").format("11.3819", l1000_gct.data_df.iloc[0, 0]))
self.assertEqual(l1000_gct.col_metadata_df.iloc[0, 0], 58,
msg=("The first value in the column metadata should be " +
"{} not {}").format("58", l1000_gct.col_metadata_df.iloc[0, 0]))
self.assertEqual(l1000_gct.row_metadata_df.iloc[0, 0], "Analyte 11",
msg=("The first value in the row metadata should be " +
"{} not {}").format("Analyte 11", l1000_gct.row_metadata_df.iloc[0, 0]))
# P100 gct
p100_file_path = os.path.join(FUNCTIONAL_TESTS_PATH, "test_p100.gct")
p100_gct = pg.parse(p100_file_path)
logger.debug("parse P100 gct - p100_gct: {}".format(p100_gct))
self.assertEqual(GCToo.GCToo, type(p100_gct))
# Check a few values
self.assertAlmostEqual(p100_gct.data_df.iloc[0, 0], 0.9182, places=4,
msg=("The first value in the data matrix should be " +
"{} not {}").format("0.9182", p100_gct.data_df.iloc[0, 0]))
self.assertEqual(p100_gct.col_metadata_df.iloc[0, 0], "MCF7",
msg=("The first value in the column metadata should be " +
"{} not {}").format("MCF7", p100_gct.col_metadata_df.iloc[0, 0]))
self.assertEqual(p100_gct.row_metadata_df.iloc[0, 0], 1859,
msg=("The first value in the row metadata should be " +
"{} not {}").format("1859", p100_gct.row_metadata_df.iloc[0, 0]))
# GCT1.2
gct_v1point2_path = os.path.join(FUNCTIONAL_TESTS_PATH, "test_v1point2_n5x10.gct")
gct_v1point2 = pg.parse(gct_v1point2_path)
logger.debug("parse GCT1.2 - gct_v1point2: {}".format(gct_v1point2))
self.assertEqual(GCToo.GCToo, type(gct_v1point2))
# Check a few values
self.assertAlmostEqual(
gct_v1point2.data_df.loc["217140_s_at", "LJP005_A375_24H_X1_B19:A06"],
6.9966, places=4)
self.assertEqual(gct_v1point2.row_metadata_df.loc["203627_at", "Description"], "IGF1R")
# Make sure col_metadata_df is empty
self.assertEqual(gct_v1point2.col_metadata_df.size, 0,
"col_metadata_df should be empty.")
#row_meta_only=True
row_metadata_df = pg.parse(l1000_file_path, row_meta_only=True)
logger.debug("row_meta_only=True - row_metadata_df.shape: {}".format(row_metadata_df.shape))
self.assertEqual(pd.DataFrame, type(row_metadata_df))
self.assertEqual((978, 11), row_metadata_df.shape)
#col_meta_only=True
col_metadata_df = pg.parse(l1000_file_path, col_meta_only=True)
logger.debug("col_meta_only=True - col_metadata_df.shape: {}".format(col_metadata_df.shape))
self.assertEqual(pd.DataFrame, type(col_metadata_df))
self.assertEqual((377, 35), col_metadata_df.shape)
#subsetting
my_rids = ["218597_s_at", "214404_x_at", "209253_at"]
my_rids_order_in_gct = ["218597_s_at", "209253_at", "214404_x_at"]
my_cidxs = [4, 0]
e_data_df = pd.DataFrame(OrderedDict([
("LJP005_A375_24H_X1_B19:A03", [10.45, 8.14, 4.92]),
("LJP005_A375_24H_X1_B19:A07", [11.04, 7.53, 6.01])
]),
dtype=np.float32)
e_data_df.index = pd.Index(my_rids_order_in_gct)
e_col_meta_df = pd.DataFrame(OrderedDict([
("pert_id", ["DMSO", "BRD-K76908866"]),
("pert_iname", ["DMSO", "CP-724714"])
]))
e_col_meta_df.index = pd.Index(
["LJP005_A375_24H_X1_B19:A03", "LJP005_A375_24H_X1_B19:A07"])
out_g = pg.parse(l1000_file_path, rid=my_rids, cidx=my_cidxs)
self.assertEqual(out_g.data_df.shape, (3, 2))
# N.B. returned object should have same order as input
pd.util.testing.assert_frame_equal(e_data_df, out_g.data_df, check_less_precise=2, check_names=False)
pd.util.testing.assert_frame_equal(e_col_meta_df, out_g.col_metadata_df[["pert_id", "pert_iname"]], check_names=False)