Version bump to 0.3, updating docs

README.rst

@@ -17,7 +17,7 @@ Metric Learning algorithms in Python.
 
 **Dependencies**
 
--  Python 2.6+
+-  Python 2.7+, 3.4+
 -  numpy, scipy, scikit-learn
 -  (for running the examples only: matplotlib)
 

doc/index.rst

@@ -17,6 +17,7 @@ metric learning algorithms.
    :caption: Algorithms
    :maxdepth: 1
 
+   metric_learn.covariance
    metric_learn.lmnn
    metric_learn.itml
    metric_learn.sdml
@@ -51,7 +52,7 @@ Alternately, download the source repository and run:
 
 **Dependencies**
 
--  Python 2.6+
+-  Python 2.7+, 3.4+
 -  numpy, scipy, scikit-learn
 -  (for running the examples only: matplotlib)
 

doc/metric_learn.covariance.rst

@@ -0,0 +1,21 @@
+Covariance metric (baseline method)
+===================================
+
+.. automodule:: metric_learn.covariance
+    :members:
+    :undoc-members:
+    :inherited-members:
+    :show-inheritance:
+
+Example Code
+------------
+
+::
+
+    from metric_learn import Covariance
+    from sklearn.datasets import load_iris
+
+    iris_data = load_iris()
+
+    cov = Covariance()
+    x = cov.fit_transform(iris_data['data'])

doc/metric_learn.itml.rst

@@ -12,19 +12,15 @@ Example Code
 
 ::
 
-    import numpy as np
-    from metric_learn import ITML
+    from metric_learn import ITML_Supervised
     from sklearn.datasets import load_iris
 
     iris_data = load_iris()
     X = iris_data['data']
     Y = iris_data['target']
 
-    itml = ITML()
-
-    num_constraints = 200
-    C = ITML.prepare_constraints(Y, X.shape[0], num_constraints)
-    itml.fit(X, C, verbose=False)
+    itml = ITML_Supervised(num_constraints=200)
+    itml.fit(X, Y)
 
 References
 ----------

doc/metric_learn.lsml.rst

@@ -12,17 +12,15 @@ Example Code
 
 ::
 
-    import numpy as np
-    from metric_learn import LSML
+    from metric_learn import LSML_Supervised
     from sklearn.datasets import load_iris
 
     iris_data = load_iris()
     X = iris_data['data']
     Y = iris_data['target']
 
-    lsml = LSML()
-    C = LSML.prepare_constraints(Y, 200)
-    isml.fit(X, C, verbose=False)
+    lsml = LSML_Supervised(num_constraints=200)
+    isml.fit(X, Y)
 
 References
 ----------

doc/metric_learn.rca.rst

@@ -12,17 +12,15 @@ Example Code
 
 ::
 
-    import numpy as np
-    from metric_learn import RCA
+    from metric_learn import RCA_Supervised
     from sklearn.datasets import load_iris
 
     iris_data = load_iris()
     X = iris_data['data']
     Y = iris_data['target']
 
-    rca = RCA()
-    C = RCA.prepare_constraints(Y, num_chunks=30, chunk_size=2)
-    rca.fit(X, C)
+    rca = RCA_Supervised(num_chunks=30, chunk_size=2)
+    rca.fit(X, Y)
 
 References
 ------------------

doc/metric_learn.sdml.rst

@@ -12,17 +12,15 @@ Example Code
 
 ::
 
-    import numpy as np
-    from metric_learn import SDML
+    from metric_learn import SDML_Supervised
     from sklearn.datasets import load_iris
 
     iris_data = load_iris()
     X = iris_data['data']
     Y = iris_data['target']
 
-    sdml = SDML()
-    W = SDML.prepare_constraints(Y, X.shape[0], 1500)
-    sdml.fit(X, W)
+    sdml = SDML_Supervised(num_constraints=200)
+    sdml.fit(X, Y)
 
 References
 ------------------

metric_learn/sdml.py

@@ -55,7 +55,8 @@ def fit(self, X, W):
     """
     X: data matrix, (n x d)
         each row corresponds to a single instance
-    W: connectivity graph, (n x n). +1 for positive pairs, -1 for negative.
+    W: connectivity graph, (n x n)
+        +1 for positive pairs, -1 for negative.
     """
     self._prepare_inputs(X, W)
     P = pinvh(self.M) + self.params['balance_param'] * self.loss_matrix

setup.py

@@ -2,7 +2,7 @@
 # -*- coding: utf-8 -*-
 from setuptools import setup
 
-version = "0.2.1"
+version = "0.3.0"
 setup(name='metric-learn',
       version=version,
       description='Python implementations of metric learning algorithms',
@@ -26,12 +26,12 @@
           'six'
       ],
       extras_require=dict(
-          docs=['sphinx', 'numpydoc'],
+          docs=['sphinx', 'shinx_rtd_theme', 'numpydoc'],
           demo=['matplotlib'],
       ),
       test_suite='test',
       keywords=[
-          'Metric Learning', 
+          'Metric Learning',
           'Large Margin Nearest Neighbor',
           'Information Theoretic Metric Learning',
           'Sparse Determinant Metric Learning',