Skip to content

Fix various bitrot issue #301

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
mgalardini merged 1 commit into from
Mar 19, 2026
Merged

Fix various bitrot issue #301

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 0 additions & 2 deletions environment.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2,13 +2,11 @@ name: pyseer_env
channels:
- conda-forge
- bioconda
- defaults
dependencies:
- python
- pip
- pytest
- numpy
- scipy
- python-dateutil >=2.5.0
- pandas
- statsmodels >=0.10
Expand Down
20 changes: 16 additions & 4 deletions pyseer/input.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -52,7 +52,7 @@ def load_phenotypes(infile, column):
p = p.dropna()

# Check type
if not np.issubdtype(p.values.dtype, np.number):
if not pd.api.types.is_numeric_dtype(p.values.dtype):
sys.stderr.write('Phenotypes must be numeric\n')
sys.exit(1)

Expand Down Expand Up @@ -113,9 +113,21 @@ def load_structure(infile, p, max_dimensions, mds_type="classic", n_cpus=1,
elif mds_type != "metric":
sys.stderr.write("Unsupported mds type chosen. Assuming metric\n")

mds = manifold.MDS(n_components=max_dimensions, metric=metric_mds, n_jobs=n_cpus,
dissimilarity='precomputed',
random_state=seed)
try:
mds = manifold.MDS(n_components=max_dimensions,
metric_mds=metric_mds,
metric='precomputed',
n_jobs=n_cpus,
normalized_stress='auto',
random_state=seed,
n_init=1)
except TypeError:
# sklearn < 1.8 does not have metric_mds
mds = manifold.MDS(n_components=max_dimensions,
metric=metric_mds,
n_jobs=n_cpus,
random_state=seed,
dissimilarity='precomputed')
projection = mds.fit_transform(m.values)

m = pd.DataFrame(projection,
Expand Down
6 changes: 3 additions & 3 deletions pyseer/model.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -131,11 +131,11 @@ def fit_null(p, m, cov, continuous, firth=False):
disp=False)
# Null fit with default optimiser may fail, Powell
# optimizer might work
except np.linalg.linalg.LinAlgError:
except np.linalg.LinAlgError:
null_res = null_mod.fit(start_params=start_vec,
method='powell',
disp=False)
except np.linalg.linalg.LinAlgError:
except np.linalg.LinAlgError:
sys.stderr.write('Matrix inversion error for null model\n')
return None
except statsmodels.tools.sm_exceptions.PerfectSeparationError:
Expand Down Expand Up @@ -346,7 +346,7 @@ def fixed_effects_regression(variant, p, k, m, c, af, pattern,
bad_chisq = True
notes.add('perfectly-separable-data')

except np.linalg.linalg.LinAlgError:
except np.linalg.LinAlgError:
# singular matrix error
bad_chisq = True
notes.add('matrix-inversion-error')
Expand Down
41 changes: 34 additions & 7 deletions tests/input_test.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -97,6 +97,12 @@ def test_load_structure(self):
0.10738983, -0.2647525 , -0.22611827, 1. , -0.7949316 ,
0.70578587, -0.43893806, -0.15941077, 0.63665674, 0.37763782,
-0.58174628, -0.19797499, 0.5000248 , -0.74400678, -0.1875957])
# scikit-learn >= 1.8.0 gives different results still
tr3 = np.array([0.33762230, 0.04617572, 0.16052974, 0.52845894, -0.90760096])
tc3 = np.array([0.33762230, -0.08747260, 0.06004778, 0.38093803, -0.25578627,
0.23132635, -1.00000000, 0.44847081, -0.43020593, 0.01492802,
0.13843412, 0.67209455, -0.40516276, 0.17481251, 0.38093804,
-0.62889757, -0.07803323, 0.46890566, -0.10716067, -0.31579913])
try:
self.assertTrue(abs((t.values[0] - tr).max()) < 1E-7)
self.assertTrue(abs((t.values[:,0] - tc).max()) < 1E-7)
Expand All @@ -106,9 +112,14 @@ def test_load_structure(self):
self.assertTrue(abs((t.values[0] - tr1).max()) < 1E-7)
self.assertTrue(abs((t.values[:,0] - tc1).max()) < 1E-7)
except AssertionError:
# scikit-learn >= 1.5.0 gives different results still
self.assertTrue(abs((t.values[0] - tr2).max()) < 1E-7)
self.assertTrue(abs((t.values[:,0] - tc2).max()) < 1E-7)
try:
# scikit-learn >= 1.5.0 gives different results still
self.assertTrue(abs((t.values[0] - tr2).max()) < 1E-7)
self.assertTrue(abs((t.values[:,0] - tc2).max()) < 1E-7)
except AssertionError:
# scikit-learn >= 1.8.0 gives different results still
self.assertTrue(abs((t.values[0] - tr3).max()) < 1E-7)
self.assertTrue(abs((t.values[:,0] - tc3).max()) < 1E-7)
t = load_structure(M, p, 5, 'metric', 1, 42)
tr = np.array([-0.97249805, -0.24747933, 0.49918088,
-0.04765291, 0.34207924])
Expand All @@ -117,12 +128,28 @@ def test_load_structure(self):
-0.34305642, 0.58112243, 0.56611122, 0.16864438,
-0.42256615, -0.30980933, -0.61344531, -0.3418638,
-0.55246211, 0.74549807, 1., 0.66051035])
self.assertTrue(abs((t.values[0] - tr).max()) < 1E-7)
self.assertTrue(abs((t.values[:,0] - tc).max()) < 1E-7)
# scikit-learn >= 1.8.0 gives different results for metric MDS
tr_metric = np.array([0.60675124, 0.24243111, 0.11760914, 0.48049612, -0.88035329])
tc_metric = np.array([0.60675124, 0.07887546, -0.91352384, -0.01243249, -0.10302767,
0.72392243, -0.43245006, -0.02775887, -1.00000000, 0.09598576,
0.14929232, 0.92884737, -0.64577922, -0.01815094, -0.01307313,
-0.75508475, 0.73135780, 0.13226520, -0.13928871, 0.61327209])
try:
self.assertTrue(abs((t.values[0] - tr).max()) < 1E-7)
self.assertTrue(abs((t.values[:,0] - tc).max()) < 1E-7)
except AssertionError:
# scikit-learn >= 1.8.0 gives different results for metric MDS
self.assertTrue(abs((t.values[0] - tr_metric).max()) < 1E-7)
self.assertTrue(abs((t.values[:,0] - tc_metric).max()) < 1E-7)
# bogus mds_type, should default to "metric"
t = load_structure(M, p, 5, 'test', 1, 42)
self.assertTrue(abs((t.values[0] - tr).max()) < 1E-7)
self.assertTrue(abs((t.values[:,0] - tc).max()) < 1E-7)
try:
self.assertTrue(abs((t.values[0] - tr).max()) < 1E-7)
self.assertTrue(abs((t.values[:,0] - tc).max()) < 1E-7)
except AssertionError:
# scikit-learn >= 1.8.0 gives different results for metric MDS
self.assertTrue(abs((t.values[0] - tr_metric).max()) < 1E-7)
self.assertTrue(abs((t.values[:,0] - tc_metric).max()) < 1E-7)
# no file to be found
with self.assertRaises(FileNotFoundError):
t = load_structure('nope', p, 5, 'test', 1, 42)
Expand Down
Loading