changed *RandomHiddenLayer to *RandomLayer, modified modules, demo and README to fit

tweaked demo notebook
changed to a more reasonable version number in the README

Author: dclambert

README.md

@@ -2,19 +2,19 @@ Python-ELM
 ==========
 
 Extreme Learning Machine implementation in Python
-Version  1.0
+Version  0.2
 
 This is an implementation of the Extreme Learning Machine in python,
 based on the scikit-learn machine learning library.
 
 Distance and dot product based hidden layers are provided via the
-RBFRandomHiddenLayer and SimpleRandomHiddenLayer classes respectively.
+RBFRandomLayer and SimpleRandomLayer classes respectively.
 
-The SimpleRandomHiddenLayer provides the following activation functions:
+The SimpleRandomLayer provides the following activation functions:
 
     tanh, sine, tribas, sigmoid, hardlim
 
-The RBFRandomHiddenLayer provides the following activation functions:
+The RBFRandomLayer provides the following activation functions:
 
     gaussian, multiquadric and polyharmonic spline ('poly_spline')
 

elm.py

@@ -31,7 +31,7 @@
 from sklearn.base import BaseEstimator, ClassifierMixin, RegressorMixin
 from sklearn.preprocessing import LabelBinarizer
 
-from random_hidden_layer import SimpleRandomHiddenLayer
+from random_layer import SimpleRandomLayer
 
 __all__ = ["ELMRegressor",
            "ELMClassifier",
@@ -104,7 +104,7 @@ class ELMRegressor(BaseELM, RegressorMixin):
     Parameters
     ----------
     `hidden_layer` : random_hidden_layer instance, optional
-        (default=SimpleRandomHiddenLayer(random_state=0))
+        (default=SimpleRandomLayer(random_state=0))
 
     `regressor`    : regressor instance, optional (default=None)
         If provided, this object is used to perform the regression from hidden
@@ -124,7 +124,7 @@ class ELMRegressor(BaseELM, RegressorMixin):
 
     See Also
     --------
-    RBFRandomHiddenLayer, SimpleRandomHiddenLayer, ELMRegressor, ELMClassifier
+    RBFRandomLayer, SimpleRandomLayer, ELMRegressor, ELMClassifier
     SimpleELMRegressor, SimpleELMClassifier
 
     References
@@ -136,7 +136,7 @@ class ELMRegressor(BaseELM, RegressorMixin):
     """
 
     def __init__(self,
-                 hidden_layer=SimpleRandomHiddenLayer(random_state=0),
+                 hidden_layer=SimpleRandomLayer(random_state=0),
                  regressor=None):
 
         super(ELMRegressor, self).__init__(hidden_layer, regressor)
@@ -231,7 +231,7 @@ class ELMClassifier(BaseELM, ClassifierMixin):
     Parameters
     ----------
     `hidden_layer` : random_hidden_layer instance, optional
-        (default=SimpleRandomHiddenLayer(random_state=0))
+        (default=SimpleRandomLayer(random_state=0))
 
     `regressor`    : regressor instance, optional (default=None)
         If provided, this object is used to perform the regression from hidden
@@ -251,7 +251,7 @@ class ELMClassifier(BaseELM, ClassifierMixin):
 
     See Also
     --------
-    RBFRandomHiddenLayer, SimpleRandomHiddenLayer, ELMRegressor, ELMClassifier
+    RBFRandomLayer, SimpleRandomLayer, ELMRegressor, ELMClassifier
     SimpleELMRegressor, SimpleELMClassifier
 
     References
@@ -262,7 +262,7 @@ class ELMClassifier(BaseELM, ClassifierMixin):
               2006.
     """
     def __init__(self,
-                 hidden_layer=SimpleRandomHiddenLayer(random_state=0),
+                 hidden_layer=SimpleRandomLayer(random_state=0),
                  regressor=None):
 
         super(ELMClassifier, self).__init__(hidden_layer, regressor)
@@ -333,7 +333,7 @@ def predict(self, X):
         return class_predictions
 
 
-# ELMRegressor with default SimpleRandomHiddenLayer
+# ELMRegressor with default SimpleRandomLayer
 class SimpleELMRegressor(BaseEstimator, RegressorMixin):
     """
     SimpleELMRegressor is a regressor based on the Extreme Learning Machine.
@@ -344,13 +344,13 @@ class SimpleELMRegressor(BaseEstimator, RegressorMixin):
     [1][2]
 
     SimpleELMRegressor is a wrapper for an ELMRegressor that uses a
-    SimpleRandomHiddenLayer and passes the __init__ parameters through
+    SimpleRandomLayer and passes the __init__ parameters through
     to the hidden layer generated by the fit() method.
 
     Parameters
     ----------
     `n_hidden` : int, optional (default=20)
-        Number of units to generate in the SimpleRandomHiddenLayer
+        Number of units to generate in the SimpleRandomLayer
 
     `activation_func` : {callable, string} optional (default='tanh')
         Function used to transform input activation
@@ -372,7 +372,7 @@ class SimpleELMRegressor(BaseEstimator, RegressorMixin):
 
     See Also
     --------
-    RBFRandomHiddenLayer, SimpleRandomHiddenLayer, ELMRegressor, ELMClassifier
+    RBFRandomLayer, SimpleRandomLayer, ELMRegressor, ELMClassifier
     SimpleELMRegressor, SimpleELMClassifier
 
     References
@@ -414,10 +414,10 @@ def fit(self, X, y):
 
             Returns an instance of self.
         """
-        rhl = SimpleRandomHiddenLayer(n_hidden=self.n_hidden,
-                                      activation_func=self.activation_func,
-                                      activation_args=self.activation_args,
-                                      random_state=self.random_state)
+        rhl = SimpleRandomLayer(n_hidden=self.n_hidden,
+                                activation_func=self.activation_func,
+                                activation_args=self.activation_args,
+                                random_state=self.random_state)
 
         self.elm_regressor_ = ELMRegressor(hidden_layer=rhl)
         self.elm_regressor_.fit(X, y)
@@ -442,7 +442,7 @@ def predict(self, X):
         return self.elm_regressor_.predict(X)
 
 
-# ELMClassifier with default SimpleRandomHiddenLayer
+# ELMClassifier with default SimpleRandomLayer
 class SimpleELMClassifier(BaseEstimator, ClassifierMixin):
     """
     SimpleELMClassifier is a classifier based on the Extreme Learning Machine.
@@ -453,13 +453,13 @@ class SimpleELMClassifier(BaseEstimator, ClassifierMixin):
     [1][2]
 
     SimpleELMClassifier is a wrapper for an ELMClassifier that uses a
-    SimpleRandomHiddenLayer and passes the __init__ parameters through
+    SimpleRandomLayer and passes the __init__ parameters through
     to the hidden layer generated by the fit() method.
 
     Parameters
     ----------
     `n_hidden` : int, optional (default=20)
-        Number of units to generate in the SimpleRandomHiddenLayer
+        Number of units to generate in the SimpleRandomLayer
 
     `activation_func` : {callable, string} optional (default='tanh')
         Function used to transform input activation
@@ -484,7 +484,7 @@ class SimpleELMClassifier(BaseEstimator, ClassifierMixin):
 
     See Also
     --------
-    RBFRandomHiddenLayer, SimpleRandomHiddenLayer, ELMRegressor, ELMClassifier
+    RBFRandomLayer, SimpleRandomLayer, ELMRegressor, ELMClassifier
     SimpleELMRegressor, SimpleELMClassifier
 
     References
@@ -547,10 +547,10 @@ def fit(self, X, y):
 
             Returns an instance of self.
         """
-        rhl = SimpleRandomHiddenLayer(n_hidden=self.n_hidden,
-                                      activation_func=self.activation_func,
-                                      activation_args=self.activation_args,
-                                      random_state=self.random_state)
+        rhl = SimpleRandomLayer(n_hidden=self.n_hidden,
+                                activation_func=self.activation_func,
+                                activation_args=self.activation_args,
+                                random_state=self.random_state)
 
         self.elm_classifier_ = ELMClassifier(hidden_layer=rhl)
         self.elm_classifier_.fit(X, y)

elm_notebook.py

@@ -14,7 +14,7 @@
 from time import time
 from sklearn.cluster import k_means
 from elm import ELMClassifier, ELMRegressor, SimpleELMClassifier, SimpleELMRegressor
-from random_hidden_layer import SimpleRandomHiddenLayer, RBFRandomHiddenLayer
+from random_layer import SimpleRandomLayer, RBFRandomLayer
 
 # <codecell>
 
@@ -64,11 +64,10 @@ def res_dist(x, y, e, n_runs=100, random_state=None):
 dgx_train, dgx_test, dgy_train, dgy_test = train_test_split(dgx, dgy, test_size=0.2)
 
 diabetes = load_diabetes()
-#dbx, dby = stdsc.fit_transform(diabetes.data), stdsc.fit_transform(diabetes.target)
 dbx, dby = stdsc.fit_transform(diabetes.data), diabetes.target
 dbx_train, dbx_test, dby_train, dby_test = train_test_split(dbx, dby, test_size=0.2)
 
-mrx, mry = make_regression(n_samples=2000)
+mrx, mry = make_regression(n_samples=2000, n_targets=2)
 mrx_train, mrx_test, mry_train, mry_test = train_test_split(mrx, mry, test_size=0.2)
 
 xtoy, ytoy = make_toy()
@@ -78,50 +77,63 @@ def res_dist(x, y, e, n_runs=100, random_state=None):
 
 # <codecell>
 
+# SimpleELMClassifier test
+elmc = SimpleELMClassifier(n_hidden=500)
+elmc.fit(dgx_train, dgy_train)
+print elmc.score(dgx_train, dgy_train), elmc.score(dgx_test, dgy_test)
+
+# <codecell>
+
+# SimpleELMRegressor test
+elmr = SimpleELMRegressor()
+elmr.fit(xtoy_train, ytoy_train)
+print elmr.score(xtoy_train, ytoy_train), elmr.score(xtoy_test, ytoy_test)
+plot(xtoy, ytoy, xtoy, elmr.predict(xtoy))
+
+# <codecell>
+
 # RBF tests
-elmc = ELMClassifier(RBFRandomHiddenLayer(activation_func='gaussian'))
+elmc = ELMClassifier(RBFRandomLayer(activation_func='gaussian'))
 tr,ts = res_dist(irx, iry, elmc, n_runs=100, random_state=0)
 
-elmc = ELMClassifier(RBFRandomHiddenLayer(activation_func='poly_spline', gamma=2))
+elmc = ELMClassifier(RBFRandomLayer(activation_func='poly_spline', gamma=2))
 tr,ts = res_dist(irx, iry, elmc, n_runs=100, random_state=0)
 
-elmc = ELMClassifier(RBFRandomHiddenLayer(activation_func='multiquadric'))
+elmc = ELMClassifier(RBFRandomLayer(activation_func='multiquadric'))
 tr,ts = res_dist(irx, iry, elmc, n_runs=100, random_state=0)
 
 # Simple tests
-elmc = ELMClassifier(SimpleRandomHiddenLayer(activation_func='sine'))
+elmc = ELMClassifier(SimpleRandomLayer(activation_func='sine'))
 tr,ts = res_dist(irx, iry, elmc, n_runs=100, random_state=0)
 
-elmc = ELMClassifier(SimpleRandomHiddenLayer(activation_func='tanh'))
+elmc = ELMClassifier(SimpleRandomLayer(activation_func='tanh'))
 tr,ts = res_dist(irx, iry, elmc, n_runs=100, random_state=0)
 
-elmc = ELMClassifier(SimpleRandomHiddenLayer(activation_func='tribas'))
+elmc = ELMClassifier(SimpleRandomLayer(activation_func='tribas'))
 tr,ts = res_dist(irx, iry, elmc, n_runs=100, random_state=0)
 
-elmc = ELMClassifier(SimpleRandomHiddenLayer(activation_func='sigmoid'))
+elmc = ELMClassifier(SimpleRandomLayer(activation_func='sigmoid'))
 tr,ts = res_dist(irx, iry, elmc, n_runs=100, random_state=0)
 
-elmc = ELMClassifier(SimpleRandomHiddenLayer(activation_func='hardlim'))
+elmc = ELMClassifier(SimpleRandomLayer(activation_func='hardlim'))
 tr,ts = res_dist(irx, iry, elmc, n_runs=100, random_state=0)
 
 # <codecell>
 
-hardlim = (lambda a: np.array(a > 0.0, dtype=float))
-tribas = (lambda a: np.clip(1.0 - np.fabs(a), 0.0, 1.0))
-elmr = ELMRegressor(SimpleRandomHiddenLayer(random_state=0, activation_func=tribas))
+elmr = ELMRegressor(SimpleRandomLayer(random_state=0, activation_func='tribas'))
 elmr.fit(xtoy_train, ytoy_train)
 print elmr.score(xtoy_train, ytoy_train), elmr.score(xtoy_test, ytoy_test)
 plot(xtoy, ytoy, xtoy, elmr.predict(xtoy))
 
 # <codecell>
 
-rhl = SimpleRandomHiddenLayer(n_hidden=200)
+rhl = SimpleRandomLayer(n_hidden=200)
 elmr = ELMRegressor(hidden_layer=rhl)
 tr, ts = res_dist(mrx, mry, elmr, n_runs=20, random_state=0)
 
 # <codecell>
 
-rhl = RBFRandomHiddenLayer(n_hidden=15, gamma=0.25)
+rhl = RBFRandomLayer(n_hidden=15, gamma=0.25)
 elmr = ELMRegressor(hidden_layer=rhl)
 elmr.fit(xtoy_train, ytoy_train)
 print elmr.score(xtoy_train, ytoy_train), elmr.score(xtoy_test, ytoy_test)
@@ -132,33 +144,33 @@ def res_dist(x, y, e, n_runs=100, random_state=None):
 nh = 10
 (ctrs, _, _) = k_means(xtoy_train, nh)
 unit_rs = np.ones(nh)
-rhl = RBFRandomHiddenLayer(n_hidden=nh, activation_func='poly_spline', gamma=3)
-#rhl = RBFRandomHiddenLayer(n_hidden=nh, activation_func='multiquadric', gamma=1)
-#rhl = RBFRandomHiddenLayer(n_hidden=nh, centers=ctrs, radii=unit_rs, gamma=4)
+#rhl = RBFRandomLayer(n_hidden=nh, activation_func='poly_spline', gamma=3)
+#rhl = RBFRandomLayer(n_hidden=nh, activation_func='multiquadric', gamma=1)
+rhl = RBFRandomLayer(n_hidden=nh, centers=ctrs, radii=unit_rs)
 elmr = ELMRegressor(hidden_layer=rhl)
 elmr.fit(xtoy_train, ytoy_train)
 print elmr.score(xtoy_train, ytoy_train), elmr.score(xtoy_test, ytoy_test)
 plot(xtoy, ytoy, xtoy, elmr.predict(xtoy))
 
 # <codecell>
 
-rbf_rhl = RBFRandomHiddenLayer(n_hidden=100, random_state=0, gamma=0.1)
+rbf_rhl = RBFRandomLayer(n_hidden=100, random_state=0, gamma=0.1)
 elmc_rbf = ELMClassifier(hidden_layer=rbf_rhl)
 elmc_rbf.fit(dgx_train, dgy_train)
 print elmc_rbf.score(dgx_train, dgy_train), elmc_rbf.score(dgx_test, dgy_test)
 
 def powtanh_xfer(activations, power=1.0):
     return pow(np.tanh(activations), power)
 
-#tanh_rhl = SimpleRandomHiddenLayer(n_hidden=5000, random_state=0)
-tanh_rhl = SimpleRandomHiddenLayer(n_hidden=5000, activation_func=powtanh_xfer, activation_args={'power':2.0})
+#tanh_rhl = SimpleRandomLayer(n_hidden=5000, random_state=0)
+tanh_rhl = SimpleRandomLayer(n_hidden=5000, activation_func=powtanh_xfer, activation_args={'power':2.0})
 elmc_tanh = ELMClassifier(hidden_layer=tanh_rhl)
 elmc_tanh.fit(dgx_train, dgy_train)
 print elmc_tanh.score(dgx_train, dgy_train), elmc_tanh.score(dgx_test, dgy_test)
 
 # <codecell>
 
-rbf_rhl = RBFRandomHiddenLayer(n_hidden=100, gamma=0.1)
+rbf_rhl = RBFRandomLayer(n_hidden=100, gamma=0.1)
 tr, ts = res_dist(dgx, dgy, ELMClassifier(hidden_layer=rbf_rhl), n_runs=100, random_state=0)
 
 # <codecell>
@@ -168,13 +180,12 @@ def powtanh_xfer(activations, power=1.0):
 
 # <codecell>
 
-from sklearn.svm import SVR
 from sklearn.ensemble import RandomForestRegressor
 tr, ts = res_dist(dbx, dby, RandomForestRegressor(n_estimators=15), n_runs=100, random_state=0)
 hist(tr), hist(ts)
 print
 
-rhl = RBFRandomHiddenLayer(n_hidden=15, gamma=0.01)
+rhl = RBFRandomLayer(n_hidden=15, gamma=0.01)
 tr,ts = res_dist(dbx, dby, ELMRegressor(rhl), n_runs=100, random_state=0)
 hist(tr), hist(ts)
 print
@@ -192,19 +203,6 @@ def powtanh_xfer(activations, power=1.0):
 
 # <codecell>
 
-elmc = SimpleELMClassifier(n_hidden=500, activation_func='hardlim')
-elmc.fit(dgx_train, dgy_train)
-print elmc.score(dgx_train, dgy_train), elmc.score(dgx_test, dgy_test)
-
-# <codecell>
-
-elmr = SimpleELMRegressor()
-elmr.fit(xtoy_train, ytoy_train)
-print elmr.score(xtoy_train, ytoy_train), elmr.score(xtoy_test, ytoy_test)
-plot(xtoy, ytoy, xtoy, elmr.predict(xtoy))
-
-# <codecell>
-
 elmr = SimpleELMRegressor(activation_func='tribas')
 elmr.fit(xtoy_train, ytoy_train)
 print elmr.score(xtoy_train, ytoy_train), elmr.score(xtoy_test, ytoy_test)