Package: release.debian.org
Severity: normal
User: release.debian....@packages.debian.org
Usertags: unblock
Please unblock package python-mdp
This version fixes FTBFS #768675. The fixes are cherry-picked from
upstream git repository. Every single commit has been incorporated
as a quilt patch. Two other bugs are fixed which were reported on
upstream user mailinglist: "Fix inclusion of CSS data in bimdp" and
"run tests from within a corresponding tests directory to work
around new py.test conftest.py autodiscovery features in python-pytest
version 2.6.3". Documentation-only fixes are also included. The
Debian changelog entry explains everything:
* New upstream git snapshot
- resolves FTBFS by a workaround of faulty unicode encoding in
sklearn docstrings (Closes: #768675). The corresponding
commits in upstream git are the quilt patches:
changeset_af5294f0b78ea5b20e4c1c23fc55a4bdaa0749c9.diff
changeset_4fc2b74375701dcabde0e3368841ce31a52c4529.diff
changeset_324cb9b963a7e9d1bb22b5934eaa9f17974f5b11.diff
- Documentation fixes. The corresponding commits in upstream
git are the quilt patches:
changeset_f4a84b7186289027410abe116f5487f800869be2.diff
changeset_4d05f0adafcc770277f36d30894a5c6aefe8a58b.diff
changeset_2b1048b980748366dd6439317fccc99f728056a7.diff
changeset_17202a65f7608e550c52953ef4026cf8fe623c16.diff
- Fix inclusion of CSS data in bimdp. The corresponding commit
in upstream git is the quilt patch:
changeset_4ec2f2940fda4f4fec9db184dbb1b93053040159.diff
* debian/rules
- run tests from within a corresponding tests directory to work around
new py.test conftest.py autodiscovery features in python-pytest
version 2.6.3
Src debdiff
diff -Nru mdp-3.3/debian/changelog mdp-3.3/debian/changelog
--- mdp-3.3/debian/changelog 2012-09-30 09:09:35.000000000 +0200
+++ mdp-3.3/debian/changelog 2014-11-28 16:45:30.000000000 +0100
@@ -1,3 +1,29 @@
+mdp (3.3-2) unstable; urgency=low
+
+ * New upstream git snapshot
+ - resolves FTBFS by a workaround of faulty unicode encoding in
+ sklearn docstrings (Closes: #768675). The corresponding
+ commits in upstream git are the quilt patches:
+ changeset_af5294f0b78ea5b20e4c1c23fc55a4bdaa0749c9.diff
+ changeset_4fc2b74375701dcabde0e3368841ce31a52c4529.diff
+ changeset_324cb9b963a7e9d1bb22b5934eaa9f17974f5b11.diff
+ - Documentation fixes. The corresponding commits in upstream
+ git are the quilt patches:
+ changeset_f4a84b7186289027410abe116f5487f800869be2.diff
+ changeset_4d05f0adafcc770277f36d30894a5c6aefe8a58b.diff
+ changeset_2b1048b980748366dd6439317fccc99f728056a7.diff
+ changeset_17202a65f7608e550c52953ef4026cf8fe623c16.diff
+ - Fix inclusion of CSS data in bimdp. The corresponding commit
+ in upstream git is the quilt patch:
+ changeset_4ec2f2940fda4f4fec9db184dbb1b93053040159.diff
+ * debian/rules
+ - run tests from within a corresponding tests directory to work around
+ new py.test conftest.py autodiscovery features in python-pytest
+ version 2.6.3
+ * Upload sponsored by Yaroslav Halchenko
+
+ -- Tiziano Zito <opossumn...@gmail.com> Wed, 12 Nov 2014 15:53:57 +0100
+
mdp (3.3-1) unstable; urgency=low
* New upstream bug-fix release
diff -Nru
mdp-3.3/debian/patches/changeset_17202a65f7608e550c52953ef4026cf8fe623c16.diff
mdp-3.3/debian/patches/changeset_17202a65f7608e550c52953ef4026cf8fe623c16.diff
---
mdp-3.3/debian/patches/changeset_17202a65f7608e550c52953ef4026cf8fe623c16.diff
1970-01-01 01:00:00.000000000 +0100
+++
mdp-3.3/debian/patches/changeset_17202a65f7608e550c52953ef4026cf8fe623c16.diff
2014-11-28 16:45:30.000000000 +0100
@@ -0,0 +1,54 @@
+From: Niko Wilbert <niko.wilb...@tngtech.com>
+Subject: DOC: some small updates and fixes in docstrings and comments
+
+Index: mdp-toolkit/bimdp/__init__.py
+===================================================================
+--- mdp-toolkit.orig/bimdp/__init__.py
++++ mdp-toolkit/bimdp/__init__.py
+@@ -42,11 +42,6 @@ This is described in more detail in the
+
+ ### T O D O ###
+
+-# ------------- optional ----------------
+-
+-# TODO: maybe also allow target==EXIT_TARGET during training
+-# would have to modify _train_node_single_phase
+-
+ # TODO: add a target seperator that does not remove the key. Could use
+ # -> remove key
+ # --> remove one '-' on entry
+@@ -70,7 +65,7 @@ This is described in more detail in the
+ # is in place.
+
+ # TODO: Implement more internal checks for node output result?
+-# Check that last element is not None? Use assume?
++# Check that last element is not None?
+
+ # TODO: implement switchlayer, a layer where each column represents a
different
+ # target, so the target value determines which nodes are used
+Index: mdp-toolkit/bimdp/binode.py
+===================================================================
+--- mdp-toolkit.orig/bimdp/binode.py
++++ mdp-toolkit/bimdp/binode.py
+@@ -60,8 +60,7 @@ keywords are treated in a special way:
+ 'target' -- If any template method like execute finds a 'target' keyword in
+ the message then this is used as the target value in the return value.
+ However, if _execute then also returns a target value this overwrites the
+- target value. In global_message calls 'target' has no special meaning and
+- can be used like any other keyword.
++ target value.
+
+ 'method' -- Specify the name of the method that should be used instead of
+ the standard one (e.g. in execute the standard method is _execute).
+Index: mdp-toolkit/bimdp/nodes/autogen.py
+===================================================================
+--- mdp-toolkit.orig/bimdp/nodes/autogen.py
++++ mdp-toolkit/bimdp/nodes/autogen.py
+@@ -1,7 +1,5 @@
+ """
+ Module to automatically create a module with BiMDP versions of MDP nodes.
+-
+-Run this module to overwrite the autogen_binodes module with a new version.
+ """
+
+ import inspect
diff -Nru
mdp-3.3/debian/patches/changeset_2b1048b980748366dd6439317fccc99f728056a7.diff
mdp-3.3/debian/patches/changeset_2b1048b980748366dd6439317fccc99f728056a7.diff
---
mdp-3.3/debian/patches/changeset_2b1048b980748366dd6439317fccc99f728056a7.diff
1970-01-01 01:00:00.000000000 +0100
+++
mdp-3.3/debian/patches/changeset_2b1048b980748366dd6439317fccc99f728056a7.diff
2014-11-28 16:45:30.000000000 +0100
@@ -0,0 +1,23 @@
+From: Alberto Escalante <alberto.nicolas.escala...@gmail.com>
+Subject: DOC: indicated what expansion failed in the tests of
GeneralExpansionNode
+
+Index: mdp-toolkit/mdp/test/test_GeneralExpansionNode.py
+===================================================================
+--- mdp-toolkit.orig/mdp/test/test_GeneralExpansionNode.py
++++ mdp-toolkit/mdp/test/test_GeneralExpansionNode.py
+@@ -21,12 +21,12 @@ def testGeneralExpansionNode():
+ input = numx.random.normal(size=(samples, input_dim))
+ out = cen.execute(input)
+ assert_array_almost_equal(out[:, 0:input_dim], input,
+- 6, "incorrect constant expansion")
++ 6, "incorrect identity expansion")
+ assert_array_almost_equal(out[:, input_dim:2*input_dim],
+- input ** 2, 6, "incorrect constant expansion")
++ input ** 2, 6, "incorrect squares expansion")
+ assert_array_almost_equal(out[:, 2*input_dim:],
+ dumb_quadratic_expansion(input), 6,
+- "incorrect constant expansion")
++ "incorrect quadratic expansion")
+ assert cen.expanded_dim(input_dim) == 2 * input_dim + input_dim**2,
"expanded_dim failed"
+ assert_array_almost_equal(cen.output_sizes(input_dim),
+ numx.array([input_dim, input_dim,
diff -Nru
mdp-3.3/debian/patches/changeset_324cb9b963a7e9d1bb22b5934eaa9f17974f5b11.diff
mdp-3.3/debian/patches/changeset_324cb9b963a7e9d1bb22b5934eaa9f17974f5b11.diff
---
mdp-3.3/debian/patches/changeset_324cb9b963a7e9d1bb22b5934eaa9f17974f5b11.diff
1970-01-01 01:00:00.000000000 +0100
+++
mdp-3.3/debian/patches/changeset_324cb9b963a7e9d1bb22b5934eaa9f17974f5b11.diff
2014-11-28 16:45:30.000000000 +0100
@@ -0,0 +1,18 @@
+From: Pietro Berkes <pber...@enthought.com>
+Subject: FIX: cStringIO does not support Unicode
+ `autogen` was breaking in nodes with Unicode characters in their docstring.
We
+traced this down to cStringIO, which unlike StringIO does not support unicode.
+
+Index: mdp-toolkit/bimdp/nodes/autogen.py
+===================================================================
+--- mdp-toolkit.orig/bimdp/nodes/autogen.py
++++ mdp-toolkit/bimdp/nodes/autogen.py
+@@ -6,7 +6,7 @@ Run this module to overwrite the autogen
+
+ import inspect
+ import mdp
+-from cStringIO import StringIO
++from StringIO import StringIO
+
+ # Blacklist of nodes that cause problems with autogeneration
+ NOAUTOGEN_MDP_NODES = [
diff -Nru
mdp-3.3/debian/patches/changeset_4d05f0adafcc770277f36d30894a5c6aefe8a58b.diff
mdp-3.3/debian/patches/changeset_4d05f0adafcc770277f36d30894a5c6aefe8a58b.diff
---
mdp-3.3/debian/patches/changeset_4d05f0adafcc770277f36d30894a5c6aefe8a58b.diff
1970-01-01 01:00:00.000000000 +0100
+++
mdp-3.3/debian/patches/changeset_4d05f0adafcc770277f36d30894a5c6aefe8a58b.diff
2014-11-28 16:45:30.000000000 +0100
@@ -0,0 +1,57 @@
+From: Alberto Escalante <alberto.nicolas.escala...@gmail.com>
+Subject: FIX: made code and documentation consistent regarding the parameter k
in invert_exp_funcs2
+
+Index: mdp-toolkit/mdp/nodes/expansion_nodes.py
+===================================================================
+--- mdp-toolkit.orig/mdp/nodes/expansion_nodes.py
++++ mdp-toolkit/mdp/nodes/expansion_nodes.py
+@@ -343,7 +343,7 @@ class GeneralExpansionNode(_ExpansionNod
+ self.input_dim,
+ self.funcs,
+ use_hint=use_hint,
+- k=0.001)
++ k=0.000001)
+ return app_x_2.astype(self.dtype)
+ except NotImplementedError, exc:
+ raise mdp.MDPException(exc)
+Index: mdp-toolkit/mdp/utils/routines.py
+===================================================================
+--- mdp-toolkit.orig/mdp/utils/routines.py
++++ mdp-toolkit/mdp/utils/routines.py
+@@ -422,19 +422,19 @@ def gabor(size, alpha, phi, freq, sgm, x
+
+ def residuals(app_x, y_noisy, exp_funcs, x_orig, k=0.0):
+ """Function used internally by invert_exp_funcs2 to approximate
+- inverses in ConstantExpansionNode. """
+- app_x = app_x.reshape((1,len(app_x)))
++ inverses in GeneralExpansionNode. """
++ app_x = app_x.reshape((1,-1))
+ app_exp_x = numx.concatenate([func(app_x) for func in exp_funcs],axis=1)
+
+ div_y = numx.sqrt(len(y_noisy))
+- div_x = numx.sqrt(len(x_orig))
+- return numx.append( (1-k)*(y_noisy-app_exp_x[0]) / div_y, k * (x_orig -
app_x[0])/div_x )
++ div_x = numx.sqrt(len(x_orig))
++ return numx.append( (1-k)**0.5 *(y_noisy-app_exp_x[0]) / div_y, (k)**0.5
* (x_orig - app_x[0])/div_x )
+
+ def invert_exp_funcs2(exp_x_noisy, dim_x, exp_funcs, use_hint=False, k=0.0):
+ """Approximates a preimage app_x of exp_x_noisy.
+
+- Returns an array app_x, such that each row of exp_x_noisy is close
+- to each row of exp_funcs(app_x).
++ Returns an array app_x, such that each row of exp_funcs(app_x) is close
++ to each row of exp_x_noisy.
+
+ use_hint: determines the starting point for the approximation of the
+ preimage. There are three possibilities.
+@@ -445,8 +445,8 @@ def invert_exp_funcs2(exp_x_noisy, dim_x
+ k: weighting factor in [0, 1] to balance between approximation error and
+ closeness to the starting point. For instance:
+ objective function is to minimize:
+- (1-k) * |exp_funcs(app_x) - exp_x_noisy|/output_dim +
+- k * |app_x - starting point|/input_dim
++ (1-k) * ||exp_funcs(app_x) - exp_x_noisy||**2/output_dim +
++ k * ||app_x - starting point||**2/input_dim
+
+ Note: this function requires scipy.
+ """
diff -Nru
mdp-3.3/debian/patches/changeset_4ec2f2940fda4f4fec9db184dbb1b93053040159.diff
mdp-3.3/debian/patches/changeset_4ec2f2940fda4f4fec9db184dbb1b93053040159.diff
---
mdp-3.3/debian/patches/changeset_4ec2f2940fda4f4fec9db184dbb1b93053040159.diff
1970-01-01 01:00:00.000000000 +0100
+++
mdp-3.3/debian/patches/changeset_4ec2f2940fda4f4fec9db184dbb1b93053040159.diff
2014-11-28 16:45:30.000000000 +0100
@@ -0,0 +1,16 @@
+From: Niko Wilbert <m...@nikowilbert.de>
+Subject: FIX: fixed broken inclusion of CSS data
+
+Index: mdp-toolkit/bimdp/hinet/bihtmlvisitor.py
+===================================================================
+--- mdp-toolkit.orig/bimdp/hinet/bihtmlvisitor.py
++++ mdp-toolkit/bimdp/hinet/bihtmlvisitor.py
+@@ -108,7 +108,7 @@ def show_biflow(flow, filename=None, tit
+ html_file = open(filename, 'w')
+ html_file.write('<html>\n<head>\n<title>%s</title>\n' % title)
+ html_file.write('<style type="text/css" media="screen">')
+- html_file.write(mdp.utils.BASIC_STYLE)
++ html_file.write(mdp.utils.basic_css())
+ html_file.write(BiHiNetHTMLVisitor.hinet_css())
+ html_file.write('</style>\n</head>\n<body>\n')
+ html_file.write('<h3>%s</h3>\n' % title)
diff -Nru
mdp-3.3/debian/patches/changeset_4fc2b74375701dcabde0e3368841ce31a52c4529.diff
mdp-3.3/debian/patches/changeset_4fc2b74375701dcabde0e3368841ce31a52c4529.diff
---
mdp-3.3/debian/patches/changeset_4fc2b74375701dcabde0e3368841ce31a52c4529.diff
1970-01-01 01:00:00.000000000 +0100
+++
mdp-3.3/debian/patches/changeset_4fc2b74375701dcabde0e3368841ce31a52c4529.diff
2014-11-28 16:45:30.000000000 +0100
@@ -0,0 +1,162 @@
+From: Niko Wilbert <niko.wilb...@tngtech.com>
+Subject: FIX: hacky fix for the unicode issue in bimdp autogen (fix for Bug
#11)
+
+Index: mdp-toolkit/bimdp/nodes/autogen.py
+===================================================================
+--- mdp-toolkit.orig/bimdp/nodes/autogen.py
++++ mdp-toolkit/bimdp/nodes/autogen.py
+@@ -14,6 +14,7 @@ NOAUTOGEN_MDP_NODES = [
+ ]
+ NOAUTOGEN_MDP_CLASSIFIERS = []
+
++
+ def _get_node_subclasses(node_class=mdp.Node, module=mdp.nodes):
+ """
+ Return all node classes in module which are subclasses of node_class.
+@@ -25,14 +26,15 @@ def _get_node_subclasses(node_class=mdp.
+ node_subclasses.append(node_subclass)
+ return node_subclasses
+
+-def _binode_code(fid, node_class, modulename, base_classname="BiNode",
++
++def _binode_code(write, node_class, modulename, base_classname="BiNode",
+ old_classname="Node"):
+ """Write code for BiMDP versions of normal node classes into module file.
+
+ It preserves the signature, which is useful for introspection (this is
+ used by the ParallelNode _default_fork implementation).
+
+- fid -- File handle of the module file.
++ write -- Function to write out code.
+ node_class -- Node class for which the new node class will be created.
+ modulename -- Name of the module where the node is from.
+ base_classname -- Base class to be used for the new nodes.
+@@ -41,11 +43,11 @@ def _binode_code(fid, node_class, module
+ """
+ node_name = node_class.__name__
+ binode_name = node_name[:-len(old_classname)] + base_classname
+- fid.write('class %s(%s, %s.%s):' %
+- (binode_name, base_classname, modulename, node_name))
++ write('class %s(%s, %s.%s):' %
++ (binode_name, base_classname, modulename, node_name))
+ docstring = ("Automatically created %s version of %s." %
+ (base_classname, node_name))
+- fid.write('\n """%s"""' % docstring)
++ write('\n """%s"""' % docstring)
+ ## define the init method explicitly to preserve the signature
+ docstring = node_class.__init__.__doc__
+ args, varargs, varkw, defaults = inspect.getargspec(node_class.__init__)
+@@ -55,36 +57,37 @@ def _binode_code(fid, node_class, module
+ if defaults is None:
+ defaults = []
+ first_default = len(args) - len(defaults)
+- fid.write('\n def __init__(self')
+- fid.write(''.join(', ' + arg for arg in args[:-len(defaults)]))
+- fid.write(''.join(', ' + arg + '=' + repr(defaults[i_arg])
+- for i_arg, arg in enumerate(args[first_default:])))
++ write('\n def __init__(self')
++ write(''.join(', ' + arg for arg in args[:-len(defaults)]))
++ write(''.join(', ' + arg + '=' + repr(defaults[i_arg])
++ for i_arg, arg in enumerate(args[first_default:])))
+ if varargs:
+- fid.write(', *%s' % varargs)
+- # always support kwargs, to prevent multiple-inheritance issues
++ write(', *%s' % varargs)
++ # always support kwargs, to prevent multiple-inheritance issues
+ if not varkw:
+ varkw = "kwargs"
+- fid.write(', **%s' % varkw)
+- fid.write('):')
++ write(', **%s' % varkw)
++ write('):')
+ if docstring:
+- fid.write('\n """%s"""' % docstring)
+- fid.write('\n super(%s, self).__init__(' % binode_name)
+- fid.write(', '.join('%s=%s' % (arg, arg) for arg in args))
++ write('\n """%s"""' % docstring)
++ write('\n super(%s, self).__init__(' % binode_name)
++ write(', '.join('%s=%s' % (arg, arg) for arg in args))
+ if varargs:
+ if args:
+- fid.write(', ')
+- fid.write('*%s' % varargs)
++ write(', ')
++ write('*%s' % varargs)
+ if args or varargs:
+- fid.write(', ')
+- fid.write('**%s' % varkw)
+- fid.write(')\n\n')
++ write(', ')
++ write('**%s' % varkw)
++ write(')\n\n')
+
+-def _binode_module(fid, node_classes, modulename="mdp.nodes",
++
++def _binode_module(write, node_classes, modulename="mdp.nodes",
+ base_classname="BiNode", old_classname="Node",
+ base_import="from bimdp import BiNode"):
+ """Write code for BiMDP versions of normal node classes into module file.
+
+- fid -- File handle of the module file.
++ write -- Function to write out code.
+ node_classes -- List of node classes for which binodes are created.
+ modulename -- Name of the module where the node is from.
+ base_classname -- Base class to be used for the new nodes.
+@@ -92,15 +95,25 @@ def _binode_module(fid, node_classes, mo
+ in the new class name.
+ base_import -- Inmport line for the base_class.
+ """
+- fid.write('"""\nAUTOMATICALLY GENERATED CODE, DO NOT MODIFY!\n\n')
+- fid.write('Edit and run autogen.py instead to overwrite this
module.\n"""')
+- fid.write('\n\nimport %s\n' % modulename)
+- fid.write(base_import + '\n\n')
++ write('"""\nAUTOMATICALLY GENERATED CODE, DO NOT MODIFY!\n\n')
++ write('Edit and run autogen.py instead to overwrite this module.\n"""')
++ write('\n\nimport %s\n' % modulename)
++ write(base_import + '\n\n')
+ for node_class in node_classes:
+- _binode_code(fid, node_class, modulename,
++ _binode_code(write, node_class, modulename,
+ base_classname=base_classname,
+ old_classname=old_classname)
+-
++
++
++def _get_unicode_write(fid):
++ def write(txt):
++ if type(txt) is unicode:
++ fid.write(txt)
++ else:
++ fid.write(unicode(txt, encoding='utf-8'))
++ return write
++
++
+ def binodes_code():
+ """Generate and import the BiNode wrappers for MDP Nodes."""
+ fid = StringIO()
+@@ -108,17 +121,23 @@ def binodes_code():
+ _get_node_subclasses(node_class=mdp.Node, module=mdp.nodes)
+ if not issubclass(node, mdp.ClassifierNode) and
+ node.__name__ not in NOAUTOGEN_MDP_NODES)
+- _binode_module(fid, nodes)
++ _binode_module(_get_unicode_write(fid), nodes)
+ return fid.getvalue()
+
+ def biclassifiers_code():
+ """Generate and import the BiClassifier wrappers for ClassifierNodes."""
+ fid = StringIO()
++ def write(txt):
++ if type(txt) is unicode:
++ fid.write(txt)
++ else:
++ fid.write(unicode(txt, encoding='utf-8'))
+ nodes = (node for node in
+ _get_node_subclasses(node_class=mdp.ClassifierNode,
+ module=mdp.nodes)
+ if node.__name__ not in NOAUTOGEN_MDP_CLASSIFIERS)
+- _binode_module(fid, nodes, base_classname="BiClassifier",
++ _binode_module(_get_unicode_write(fid), nodes,
++ base_classname="BiClassifier",
+ old_classname="Classifier",
+ base_import="from bimdp import BiClassifier")
+ return fid.getvalue()
diff -Nru
mdp-3.3/debian/patches/changeset_af5294f0b78ea5b20e4c1c23fc55a4bdaa0749c9.diff
mdp-3.3/debian/patches/changeset_af5294f0b78ea5b20e4c1c23fc55a4bdaa0749c9.diff
---
mdp-3.3/debian/patches/changeset_af5294f0b78ea5b20e4c1c23fc55a4bdaa0749c9.diff
1970-01-01 01:00:00.000000000 +0100
+++
mdp-3.3/debian/patches/changeset_af5294f0b78ea5b20e4c1c23fc55a4bdaa0749c9.diff
2014-11-28 16:45:30.000000000 +0100
@@ -0,0 +1,26 @@
+From: Tiziano Zito <opossumn...@gmail.com>
+Subject: FIX: wrap sklearn algs in modules ending with "_"
+ For historical reasons we were not wrapping sklearn algs in modules with
names
+ending with "_". This restriction prevented us from wrapping algs that we could
+have wrapped. We lift it now, since it does not seem to have any bad side
+effects anymore. A proper validation would be to really test that we can run
the
+algs.
+
+Index: mdp-toolkit/mdp/nodes/scikits_nodes.py
+===================================================================
+--- mdp-toolkit.orig/mdp/nodes/scikits_nodes.py
++++ mdp-toolkit/mdp/nodes/scikits_nodes.py
+@@ -174,10 +174,9 @@ def apply_to_scikits_algorithms(current_
+ if (inspect.isclass(member) and
+ member not in processed_classes):
+ processed_classes.append(member)
+- if ((hasattr(member, 'fit')
+- or hasattr(member, 'predict')
+- or hasattr(member, 'transform'))
+- and not member.__module__.endswith('_')):
++ if (hasattr(member, 'fit')
++ or hasattr(member, 'predict')
++ or hasattr(member, 'transform')):
+ action(member)
+
+ # other modules
diff -Nru
mdp-3.3/debian/patches/changeset_f4a84b7186289027410abe116f5487f800869be2.diff
mdp-3.3/debian/patches/changeset_f4a84b7186289027410abe116f5487f800869be2.diff
---
mdp-3.3/debian/patches/changeset_f4a84b7186289027410abe116f5487f800869be2.diff
1970-01-01 01:00:00.000000000 +0100
+++
mdp-3.3/debian/patches/changeset_f4a84b7186289027410abe116f5487f800869be2.diff
2014-11-28 16:45:30.000000000 +0100
@@ -0,0 +1,85 @@
+From: Alberto Escalante <alberto.nicolas.escala...@gmail.com>
+Subject: DOC: added an example to GeneralExpansionNode and made the
documentation more explicit
+
+Index: mdp-toolkit/mdp/nodes/expansion_nodes.py
+===================================================================
+--- mdp-toolkit.orig/mdp/nodes/expansion_nodes.py
++++ mdp-toolkit/mdp/nodes/expansion_nodes.py
+@@ -283,14 +283,38 @@ class GrowingNeuralGasExpansionNode(Grow
+
+
+ class GeneralExpansionNode(_ExpansionNode):
+- """Expands the input signal x according to a list [f_0, ... f_k]
+- of functions.
+-
+- Each function f_i should take the whole two-dimensional array x as input
and
+- output another two-dimensional array. Moreover the output dimension should
+- depend only on the input dimension.
+- The output of the node is [f_0[x], ... f_k[x]], that is, the concatenation
+- of each one of the outputs f_i[x].
++ """Expands the input samples by applying to them one or more functions
provided.
++
++ The functions to be applied are specified by a list [f_0, ..., f_k],
where
++ f_i, for 0 <= i <= k, denotes a particular function.
++ The input data given to these functions is a two-dimensional array and
++ the output is another two-dimensional array. The dimensionality of the
output
++ should depend only on the dimensionality of the input.
++ Given a two-dimensional input array x, the output of the node
++ is then [f_0(x), ..., f_k(x)], that is, the concatenation of each one of
++ the computed arrays f_i(x).
++
++ This node has been designed to facilitate nonlinear, fixed but arbitrary
++ transformations of the data samples within MDP flows.
++
++ **Example**::
++
++ >>> import mdp
++ >>> from mdp import numx
++
++ >>> def identity(x): return x
++
++ >>> def u3(x): return numx.absolute(x)**3 #A simple nonlinear
transformation
++
++ >>> def norm2(x): #Computes the norm of each sample returning an Nx1
array
++ >>> return ((x**2).sum(axis=1)**0.5).reshape((-1,1))
++
++ >>> x = numx.array([[-2., 2.], [0.2, 0.3], [0.6, 1.2]])
++ >>> gen = mdp.nodes.GeneralExpansionNode(funcs=[identity, u3, norm2])
++ >>> print(gen.execute(x))
++ >>> [[-2. 2. 8. 8. 2.82842712]
++ >>> [ 0.2 0.3 0.008 0.027 0.36055513]
++ >>> [ 0.6 1.2 0.216 1.728 1.34164079]]
+
+ Original code contributed by Alberto Escalante.
+ """
+@@ -299,7 +323,7 @@ class GeneralExpansionNode(_ExpansionNod
+ Short argument description:
+
+ ``funcs``
+- list of functions f_i that realize the expansion
++ list of functions f_i that realize the expansion.
+ """
+ self.funcs = funcs
+ super(GeneralExpansionNode, self).__init__(input_dim, dtype)
+@@ -312,7 +336,7 @@ class GeneralExpansionNode(_ExpansionNod
+
+ def output_sizes(self, n):
+ """Return the individual output sizes of each expansion function
+- when the input has lenght n"""
++ when the input has lenght n."""
+ sizes = numx.zeros(len(self.funcs))
+ x = numx.zeros((1,n))
+ for i, func in enumerate(self.funcs):
+@@ -333,8 +357,10 @@ class GeneralExpansionNode(_ExpansionNod
+ scipy.optimize.
+
+ ``use_hint``
+- when calculating the pseudo inverse of the expansion,
+- the hint determines the starting point for the approximation
++ when calculating a pseudo inverse of the expansion,
++ the hint determines the starting point for the approximation.
++ For details on this parameter see the function
++ ``invert_exp_funcs2`` in ``mdp.utils.routines.py``.
+
+ This method requires scipy."""
+
diff -Nru mdp-3.3/debian/patches/series mdp-3.3/debian/patches/series
--- mdp-3.3/debian/patches/series 2012-09-28 17:26:53.000000000 +0200
+++ mdp-3.3/debian/patches/series 2014-11-28 16:45:30.000000000 +0100
@@ -1 +1,9 @@
+changeset_4ec2f2940fda4f4fec9db184dbb1b93053040159.diff
+changeset_4fc2b74375701dcabde0e3368841ce31a52c4529.diff
+changeset_17202a65f7608e550c52953ef4026cf8fe623c16.diff
+changeset_324cb9b963a7e9d1bb22b5934eaa9f17974f5b11.diff
+changeset_af5294f0b78ea5b20e4c1c23fc55a4bdaa0749c9.diff
+changeset_4d05f0adafcc770277f36d30894a5c6aefe8a58b.diff
+changeset_2b1048b980748366dd6439317fccc99f728056a7.diff
+changeset_f4a84b7186289027410abe116f5487f800869be2.diff
diff -Nru mdp-3.3/debian/rules mdp-3.3/debian/rules
--- mdp-3.3/debian/rules 2012-09-28 17:26:53.000000000 +0200
+++ mdp-3.3/debian/rules 2014-11-28 16:45:30.000000000 +0100
@@ -34,7 +34,8 @@
ifeq (,$(filter nocheck,$(DEB_BUILD_OPTIONS)))
set -xe && for v in $(PYTHONALL); do \
LOC=$(INSTALLPATH)$(call py_libdir_sh, $$v); \
- python$$v $$LOC/mdp/test/run_tests.py --seed $(RSEED) $$LOC/mdp
$$LOC/bimdp; \
+ cd $$LOC/mdp/test && python$$v $$LOC/mdp/test/run_tests.py --seed
$(RSEED) && \
+ cd $$LOC/bimdp/test && python$$v $$LOC/mdp/test/run_tests.py --seed
$(RSEED) ; \
done
: # Prune compiled code which could have been generated during testing
find \( -name __pycache__ -o -name \*.pyc -o -name \*.egg-info \)
-delete
unblock python-mdp/3.3-2
-- System Information:
Debian Release: 8.0
APT prefers unstable
APT policy: (500, 'unstable'), (101, 'experimental')
Architecture: amd64 (x86_64)
Foreign Architectures: i386
Kernel: Linux 3.17-1-amd64 (SMP w/8 CPU cores)
Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
