Merge

epydoc

@@ -7,7 +7,7 @@
 # The list of objects to document.  Objects can be named using
 # dotted names, module filenames, or package directory names.
 # Alases for this option include "objects" and "values".
-modules: *.py, gof/*.py, scipy, numpy
+modules: *.py, gof/*.py, joseph/*.py, tlearn/*.py, scipy, numpy
 
 # The type of output that should be generated.  Should be one
 # of: html, text, latex, dvi, ps, pdf.
@@ -135,7 +135,7 @@ dotpath: /usr/bin/dot
 
 # The name of one or more pstat files (generated by the profile
 # or hotshot module).  These are used to generate call graphs.
-#pstat: sparse.pstat
+pstat: autotest.pstat
 
 # Specify the font used to generate Graphviz graphs.
 # (e.g., helvetica or times).

epydoc-fast

@@ -7,7 +7,7 @@
 # The list of objects to document.  Objects can be named using
 # dotted names, module filenames, or package directory names.
 # Alases for this option include "objects" and "values".
-modules: *.py, gof/*.py
+modules: *.py, gof/*.py, joseph/*.py, tlearn/*.py
 
 # The type of output that should be generated.  Should be one
 # of: html, text, latex, dvi, ps, pdf.
@@ -135,7 +135,7 @@ dotpath: /usr/bin/dot
 
 # The name of one or more pstat files (generated by the profile
 # or hotshot module).  These are used to generate call graphs.
-#pstat: sparse.pstat
+pstat: autotest.pstat
 
 # Specify the font used to generate Graphviz graphs.
 # (e.g., helvetica or times).

epydoc-fast.sh

 #!/bin/sh
 
+# python -m profile  -o autotest.pstat autotest.py
 epydoc --config epydoc-fast

epydoc.sh

 #!/bin/sh
 
+python -m profile  -o autotest.pstat autotest.py
 epydoc --config epydoc

gof/op.py

@@ -17,7 +17,7 @@ __all__ = ['Op',
 
 
 def constructor(op_cls):
-    """Make an Op look like a Result-valued function."""
+    """Make an Op look like a L{Result}-valued function."""
     def f(*args, **kwargs):
         op = op_cls(*args, **kwargs)
         if len(op.outputs) > 1:
@@ -28,12 +28,12 @@ def constructor(op_cls):
 
 class Op(object):
     """
-    Op represents a computation on the storage in its 'inputs' slot,
-    the results of which are stored in the Result instances in the
-    'outputs' slot. The owner of each Result in the outputs list must
-    be set to this Op and thus any Result instance is in the outputs
-    list of at most one Op, its owner. It is the responsibility of the
-    Op to ensure that it owns its outputs and it is encouraged (though
+    L{Op} represents a computation on the storage in its 'inputs' slot,
+    the results of which are stored in the L{Result} instances in the
+    'outputs' slot. The owner of each L{Result} in the outputs list must
+    be set to this L{Op} and thus any L{Result} instance is in the outputs
+    list of at most one L{Op}, its owner. It is the responsibility of the
+    L{Op} to ensure that it owns its outputs and it is encouraged (though
     not required) that it creates them.
     """
 

sparse.py

@@ -9,7 +9,12 @@ import tensor
 # Wrapper type
 
 def assparse(sp, **kwargs):
-    """Return SparseR version of sp"""
+    """
+    Wrapper around SparseR constructor.
+    @param sp:  A sparse matrix. assparse reads dtype and format properties
+                out of this sparse matrix.
+    @return:    SparseR version of sp.
+    """
     if isinstance(sp, SparseR):
         return sp
     else:
@@ -37,6 +42,13 @@ class SparseR(gof.result.ResultBase):
     dtype_set = set(['int', 'int32', 'int64', 'float32', 'float64'])
 
     def __init__(self, dtype, format, **kwargs):
+        """
+        Fundamental way to do create a sparse node.
+        @param dtype:   Type of numbers in the matrix.
+        @param format:  The sparse storage strategy.
+        @return         An empty SparseR instance.
+        """
+
         gof.ResultBase.__init__(self, **kwargs)
         if dtype in SparseR.dtype_set:
             self._dtype = dtype
@@ -165,8 +177,10 @@ if 0:
         def gen_outputs(self): return [SparseR()]
         def impl(x,y):
             if hasattr(x, 'getnnz'):
+                # if x is sparse, then do this.
                 return x.dot(y)
             else:
+                # if x is dense (and y is sparse), we do this
                 return y.transpose().dot(x.transpose()).transpose()
 
         def grad(self, x, y, gz):