fixed t-test and rank test

Author: davidsebfischer

diffxpy/testing/det.py

@@ -14,7 +14,7 @@
 except ImportError:
     anndata = None
 
-from batchglm.models.base import _EstimatorBase
+from batchglm.models.base import _EstimatorBase, _InputDataBase
 
 from ..stats import stats
 from . import correction
@@ -1518,6 +1518,10 @@ def __init__(
             is_sig_zerovar: bool = True
     ):
         super().__init__()
+        if isinstance(data, anndata.AnnData) or isinstance(data, anndata.Raw):
+            data = data.X
+        elif isinstance(data, _InputDataBase):
+            data = data.x
         self._x = data
         self.sample_description = sample_description
         self.grouping = grouping
@@ -1652,6 +1656,10 @@ def __init__(
             is_sig_zerovar: bool = True
     ):
         super().__init__()
+        if isinstance(data, anndata.AnnData) or isinstance(data, anndata.Raw):
+            data = data.X
+        elif isinstance(data, _InputDataBase):
+            data = data.x
         self._x = data
         self.sample_description = sample_description
         self.grouping = grouping
@@ -1680,8 +1688,8 @@ def __init__(
         # TODO: can this be done on sparse?
         pval = np.zeros([data.shape[1]]) + np.nan
         pval[idx_run] = stats.mann_whitney_u_test(
-            x0=x0.x[:, idx_run].toarray(),
-            x1=x1.x[:, idx_run].toarray()
+            x0=np.asarray(x0[:, idx_run]),
+            x1=np.asarray(x1[:, idx_run])
         )
         pval[np.where(np.logical_and(
             np.logical_and(mean_x0 == mean_x1, self._mean > 0),

diffxpy/testing/tests.py

@@ -209,7 +209,7 @@ def lrt(
         noise_model="nb",
         size_factors: Union[np.ndarray, pd.core.series.Series, np.ndarray] = None,
         batch_size: int = None,
-        training_strategy: Union[str, List[Dict[str, object]], Callable] = "DEFAULT",
+        training_strategy: Union[str, List[Dict[str, object]], Callable] = "AUTO",
         quick_scale: bool = False,
         dtype="float64",
         **kwargs

diffxpy/testing/utils.py

@@ -103,6 +103,7 @@ def parse_grouping(data, sample_description, grouping):
 
 
 def split_x(data, grouping):
+    grouping = np.asarray(grouping)
     groups = np.unique(grouping)
     x0 = data[np.where(grouping == groups[0])[0]]
     x1 = data[np.where(grouping == groups[1])[0]]

diffxpy/unit_test/test_single_de.py

@@ -5,7 +5,7 @@
 import diffxpy.api as de
 
 
-class _TestSingleDE:
+class _TestSingleDe:
 
     def _prepare_data(
             self,
@@ -109,7 +109,7 @@ def _test_t_test_de(
         )
 
         test = de.test.t_test(
-            data=sim.x,
+            data=sim.input_data,
             grouping="condition",
             sample_description=sim.sample_description
         )
@@ -191,7 +191,7 @@ def _test_lrt_de(
         return True
 
 
-class TestSingleDE_STANDARD(_TestSingleDE, unittest.TestCase):
+class TestSingleDeStandard(_TestSingleDe, unittest.TestCase):
     """
     Noise model-independent tests unit tests that tests false positive and false negative rates.
     """
@@ -235,7 +235,7 @@ def test_rank_de(
         )
 
 
-class TestSingleDE_NB(_TestSingleDE, unittest.TestCase):
+class TestSingleDeNb(_TestSingleDe, unittest.TestCase):
     """
     Negative binomial noise model unit tests that tests false positive and false negative rates.
     """
@@ -281,7 +281,7 @@ def test_lrt_de_nb(
         )
 
 
-class TestSingleDE_NORM(_TestSingleDE, unittest.TestCase):
+class TestSingleDeNorm(_TestSingleDe, unittest.TestCase):
     """
     Normal noise model unit tests that tests false positive and false negative rates.
     """

diffxpy/unit_test/test_single_null.py

@@ -101,7 +101,7 @@ def _test_null_distribution_wald_multi(
             training_strategy="DEFAULT",
             dtype="float64"
         )
-        summary = test.summary()
+        _ = test.summary()
 
         # Compare p-value distribution under null model against uniform distribution.
         pval_h0 = stats.kstest(test.pval, 'uniform').pvalue
@@ -248,6 +248,7 @@ class TestSingleNullStandard(_TestSingleNull, unittest.TestCase):
     Noise model-independent unit tests that test whether a test generates uniformly
     distributed p-values if data are sampled from the null model.
     """
+
     def test_null_distribution_ttest(
             self,
             n_cells: int = 2000,
@@ -291,7 +292,7 @@ def test_null_distribution_rank(
         )
 
 
-class TestSingleNullNB(_TestSingleNull, unittest.TestCase):
+class TestSingleNullNb(_TestSingleNull, unittest.TestCase):
     """
     Negative binomial noise model unit tests that test whether a test generates uniformly
     distributed p-values if data are sampled from the null model.
@@ -365,7 +366,7 @@ def test_null_distribution_lrt_nb(
         )
 
 
-class TestSingleNullNORM(_TestSingleNull, unittest.TestCase):
+class TestSingleNullNorm(_TestSingleNull, unittest.TestCase):
     """
     Normal noise model unit tests that test whether a test generates uniformly
     distributed p-values if data are sampled from the null model.