remove random_forest module and add kmerf to independence

hyppo/__init__.py

@@ -3,6 +3,5 @@
 import hyppo.time_series
 import hyppo.tools
 import hyppo.discrim
-import hyppo.random_forest
 
 __version__ = "0.1.3"

hyppo/independence/__init__.py

@@ -4,5 +4,6 @@
 from .dcorr import Dcorr
 from .hsic import Hsic
 from .mgc import MGC
+from .kmerf import KMERF
 
-__all__ = ["RV", "CCA", "HHG", "Dcorr", "Hsic", "MGC"]
+__all__ = [s for s in dir()] # add imported tests to __all__

hyppo/independence/_utils.py

@@ -69,3 +69,15 @@ def _check_min_samples(self):
  def _check_variance(self):
  if np.var(self.x) == 0 or np.var(self.y) == 0:
  raise ValueError("Test cannot be run, one of the inputs has 0 variance")
+
+
+def sim_matrix(model, x):
+ terminals = model.apply(x)
+ ntrees = terminals.shape[1]
+
+ proxMat = 1 * np.equal.outer(terminals[:, 0], terminals[:, 0])
+ for i in range(1, ntrees):
+ proxMat += 1 * np.equal.outer(terminals[:, i], terminals[:, i])
+ proxMat = proxMat / ntrees
+
+ return proxMat

hyppo/random_forest/kmerf.py → hyppo/independence/kmerf.py

@@ -2,9 +2,9 @@
 from sklearn.ensemble import RandomForestRegressor, RandomForestClassifier
 from sklearn.metrics import pairwise_distances
 
-from .base import RandomForestTest
+from .base import IndependenceTest
 from ._utils import _CheckInputs, sim_matrix
-from ..independence import Dcorr
+from . import Dcorr
 from ..tools import perm_test
 
 
@@ -14,7 +14,7 @@
 }
 
 
-class KMERF(RandomForestTest):
+class KMERF(IndependenceTest):
  r"""
  Class for calculating the random forest based Dcorr test statistic and p-value.
  """
@@ -24,7 +24,7 @@ def __init__(self, forest="regressor", ntrees=500, **kwargs):
  self.clf = FOREST_TYPES[forest](n_estimators=ntrees, **kwargs)
  else:
  raise ValueError("Forest must be of type classification or regression")
- RandomForestTest.__init__(self)
+ IndependenceTest.__init__(self)
 
  def _statistic(self, x, y):
  r"""