slightly better handling of model args

epistasis/models/__init__.py

@@ -2,13 +2,21 @@
 A library of models to decompose high-order epistasis in genotype-phenotype
 maps.
 """
+# Import linear models
 from .linear import (EpistasisLinearRegression,
                      EpistasisLasso,
                      EpistasisRidge,
                      EpistasisElasticNet)
+
+# Import nonlinear models
 from .nonlinear import (EpistasisNonlinearRegression,
                         EpistasisPowerTransform,
                         EpistasisSpline)
+
+# Import classifiers
 from .classifiers import (EpistasisLogisticRegression,
-                          EpistasisGaussianMixture)
+                          EpistasisGaussianMixture,
+                          EpistasisGaussianProcess)
+
+# Import Pipeline object fro stitching models.
 from .pipeline import EpistasisPipeline

epistasis/models/base.py

@@ -435,9 +435,13 @@ def gpm(self):
 
     def _X(self, data=None, method=None):
         """Handle the X argument in this model."""
+        # Get object type.
+        obj = data.__class__
+
         X = data
         # If X is None, see if we saved an array.
         if X is None:
+
             # Get X from genotypes
             X = genotypes_to_X(
                 self.gpm.wildtype,
@@ -447,7 +451,8 @@ def _X(self, data=None, method=None):
                 model_type=self.model_type
             )
 
-        elif isinstance(X, str) and X in self.gpm.genotypes:
+        elif obj is str and X in self.gpm.genotypes:
+
             # Get X from genotypes
             X = genotypes_to_X(
                 self.gpm.wildtype,
@@ -458,15 +463,15 @@ def _X(self, data=None, method=None):
             )
 
         # If X is a keyword in Xbuilt, use it.
-        elif isinstance(X, str) and X in self.Xbuilt:
+        elif obj is str and X in self.Xbuilt:
             X = self.Xbuilt[X]
 
         # If 2-d array, keep as so.
-        elif isinstance(X, np.ndarray) and X.ndim == 2:
+        elif obj is np.ndarray and X.ndim == 2:
             pass
 
         # If list of genotypes.
-        elif isinstance(X, list) or isinstance(X, np.ndarray):
+        elif obj in [list, np.ndarray, pd.DataFrame, pd.Series]:
             # Get X from genotypes
             X = genotypes_to_X(
                 self.gpm.wildtype,
@@ -484,44 +489,53 @@ def _X(self, data=None, method=None):
 
     def _y(self, data=None, method=None):
         """Handle y arguments in this model."""
+        # Get object type.
+        obj = data.__class__
         y = data
+
         if y is None:
             return self.gpm.phenotypes
 
-        elif isinstance(y, np.ndarray) or isinstance(y, list):
+        elif obj in [list, np.ndarray, pd.Series, pd.DataFrame]:
             return y
 
         else:
             raise Exception("y is invalid.")
 
     def _yerr(self, data=None, method=None):
         """Handle yerr argument in this model."""
+        # Get object type.
+        obj = data.__class__
         yerr = data
         if yerr is None:
             return self.gpm.std.upper
 
-        elif isinstance(yerr, np.ndarray) or isinstance(yerr, list):
+        elif obj in [list, np.ndarray, pd.Series, pd.DataFrame]:
             return yerr
         else:
             raise Exception("yerr is invalid.")
 
     def _thetas(self, data=None, method=None):
         """Handle yerr argument in this model."""
+        # Get object type.
+        obj = data.__class__
         thetas = data
         if thetas is None:
             return self.thetas
 
-        elif isinstance(thetas, np.ndarray) or isinstance(thetas, list):
+        elif obj in [list, np.ndarray, pd.Series, pd.DataFrame]:
             return thetas
         else:
             raise Exception("thetas is invalid.")
 
     def _lnprior(self, data=None, method=None):
+        # Get object type.
+        obj = data.__class__
         _lnprior = data
         if _lnprior is None:
             return np.zeros(self.gpm.n)
 
-        elif isinstance(_lnprior, np.ndarray) or isinstance(_lnprior, list):
+        elif obj in [list, np.ndarray, pd.Series, pd.DataFrame]:
             return _lnprior
         else:
             raise Exceptison("_prior is invalid.")

epistasis/models/classifiers/__init__.py

@@ -1,2 +1,3 @@
 from .logistic import EpistasisLogisticRegression
 from .gmm import EpistasisGaussianMixture
+from .gaussian_process import EpistasisGaussianProcess