Merge pull request #85 from theislab/dev

Dev

Author: davidsebfischer

batchglm/data.py

@@ -217,15 +217,16 @@ def constraint_system_from_star(
             sample_description=sample_description,
             formula=formula,
             as_categorical=as_categorical,
-            constraints=constraints
+            constraints=constraints,
+            return_type="dataframe"
         )
     elif isinstance(constraints, tuple) or isinstance(constraints, list):
         cmat = constraint_matrix_from_string(
             dmat=dmat,
             constraints=constraints
         )
     elif isinstance(constraints, np.ndarray):
-        cmat = parse_constraints
+        cmat = constraints
     elif constraints is None:
         cmat = None
     else:
@@ -238,7 +239,8 @@ def constraint_matrix_from_dict(
         sample_description: pd.DataFrame,
         formula: str,
         as_categorical: Union[bool, list] = True,
-        constraints: dict = {}
+        constraints: dict = {},
+        return_type: str = "dataframe"
 ) -> Tuple:
     """
     Create a design matrix from some sample description and a constraint matrix
@@ -303,9 +305,14 @@ def constraint_matrix_from_dict(
     # Build constraint matrix.
     constraints_ar = constraint_matrix_from_string(
         dmat=dmat,
+        coef_names=coef_names,
         constraints=constraints_ls
     )
 
+    # Format return type
+    if return_type == "dataframe":
+        dmat = pd.DataFrame(dmat, columns=coef_names)
+
     return dmat, constraints_ar
 
 
@@ -362,6 +369,7 @@ def string_constraints_from_dict(
 
 def constraint_matrix_from_string(
         dmat: np.ndarray,
+        coef_names: list,
         constraints: Union[Tuple[str, str], List[str]]
 ):
     r"""
@@ -375,10 +383,10 @@ def constraint_matrix_from_string(
     """
     assert len(constraints) > 0, "supply constraints"
 
-    n_par_all = dmat.values.shape[1]
+    n_par_all = dmat.shape[1]
     n_par_free = n_par_all - len(constraints)
 
-    di = patsy.DesignInfo(dmat.coords["design_params"].values)
+    di = patsy.DesignInfo(coef_names)
     constraint_ls = [di.linear_constraint(x).coefs[0] for x in constraints]
     idx_constr = np.asarray([np.where(x == 1)[0][0] for x in constraint_ls])
     idx_depending = [np.where(x == 1)[0][1:] for x in constraint_ls]

batchglm/models/base_glm/utils.py

@@ -137,7 +137,7 @@ def closedform_glm_scale(
     :return: tuple (groupwise_scales, logphi, rmsd)
     """
     if size_factors is not None:
-        x = np.divide(x, size_factors)
+        x = x / size_factors
 
     # to circumvent nonlocal error
     provided_groupwise_means = groupwise_means

batchglm/unit_test/test_acc_glm_all.py

@@ -306,13 +306,6 @@ def basic_test(
         else:
             raise ValueError("noise model %s not recognized" % self.noise_model)
 
-        estimator = _TestAccuracyGlmAllEstim(
-            simulator=self.simulator(train_loc=train_loc),
-            quick_scale=False if train_scale else True,
-            noise_model=self.noise_model,
-            sparse=sparse,
-            init_mode=init_mode
-        )
         for algo in algos:
             logger.info("algorithm: %s" % algo)
             if algo in ["ADAM", "RMSPROP", "GD"]:
@@ -348,6 +341,13 @@ def basic_test(
                 glm.pkg_constants.JACOBIAN_MODE = "analytic"
             else:
                 return ValueError("algo %s not recognized" % algo)
+            estimator = _TestAccuracyGlmAllEstim(
+                simulator=self.simulator(train_loc=train_loc),
+                quick_scale=False if train_scale else True,
+                noise_model=self.noise_model,
+                sparse=sparse,
+                init_mode=init_mode
+            )
             estimator.estimate(
                 algo=algo,
                 batched=batched,