Merge pull request #550 from lamblin/win64_install

NEWS.txt

@@ -7,6 +7,7 @@ Documentation
    This explains how to make a Theano Op from a Python function.
    http://deeplearning.net/software/theano/tutorial/extending_theano.html
    (Frédéric B.)
+ * New installation instructions for Windows using EPD (Pascal L.)
 
 Interface changes
  * In 0.5, we removed the deprecated sharedvar.value property.
@@ -46,6 +47,8 @@ Sparse Sandbox Additions (not reviewed/documented/tested, but used by some peopl
 Internal changes
  * Define new exceptions MissingInputError and UnusedInputError, and use them
    in theano.function, instead of TypeError and ValueError. (Pascal L.)
+ * Better handling of bitwidth and max values of integers and pointers
+   across platforms (Pascal L.)
 
 Crash Fix
  * Do not try to use the BLAS library when blas.ldflags is manually set to an
@@ -62,6 +65,8 @@ Crash Fix
    (Frederic B., reported by Sander Dieleman)
  * GpuSoftmaxWithBias with shape (0, N) with N > 1.
    (Frédéric B., reported by Razvan P.)
+ * Fix crash under 64-bit Windows, when taking subtensors of the form a[n:]
+   (Pascal L., reported by Simon McGregor)
 
 =============
 Release Notes

doc/extending/op.txt

@@ -289,6 +289,7 @@ node representing the application of Op ``mul`` to inputs ``x`` and
 ``y``, giving a Variable instance of type ``double`` as the output.
 
 .. note::
+
    Theano relies on the fact that if you call the ``make_node`` method
    of Apply's first argument on the inputs passed as the Apply's
    second argument, the call will not fail and the returned Apply
@@ -315,6 +316,7 @@ contain a single storage cell for the multiplication's variable.
 Here, ``z`` is a list of one element. By default, ``z == [None]``.
 
 .. note::
+
    It is possible that ``z`` does not contain ``None``. If it contains
    anything else, Theano guarantees that whatever it contains is what
    ``perform`` put there the last time it was called with this
@@ -324,6 +326,7 @@ Here, ``z`` is a list of one element. By default, ``z == [None]``.
    :ref:`op` documentation.
 
 .. warning::
+
    We gave ``z`` the Theano type ``double`` in ``make_node``, which means
    that a Python ``float`` must be put there. You should not put, say, an
    ``int`` in ``z[0]`` because Theano assumes Ops handle typing properly.

doc/extending/optimization.txt

@@ -232,6 +232,7 @@ arithmetics that your Ops implement. Theano might provide facilities
 for this somewhere in the future.
 
 .. note::
+
    :class:`Env` is a Theano structure intended for the optimization
    phase. It is used internally by function and Module and is rarely
    exposed to the end user. You can use it to test out optimizations,

doc/extending/pipeline.txt

@@ -98,6 +98,7 @@ case if ``borrow`` was True, the thunk would be allowed to reuse (or
 "recycle") the storage).
 
 .. note::
+
     Compiled libraries are stored within a specific compilation directory,
     which by default is set to ``$HOME/.theano/compiledir_xxx``, where
     ``xxx`` identifies the platform (under Windows the default location

doc/install.txt

@@ -600,7 +600,51 @@ Windows
 Installing Dependencies
 ~~~~~~~~~~~~~~~~~~~~~~~
 
-If neither Python nor MinGW is installed on your computer, the easiest way to
+In academia: EPD
+################
+
+If you are working in academia, the easiest way to install most of the
+dependencies is to install `Enthought Python Distribution (EPD) <epd>`_.
+If you are affiliated to a university (as student or employee), you can
+download the installation for free.
+
+EPD installation includes in particular Python (and the development headers),
+numpy, scipy, nose, sphinx, easy_install, pydot (but *not* `Graphviz`_, which is
+necessary for it to work), g++, and the MKL
+implementation of blas.
+
+If you want to use the iPython shell, you should first try to import numpy
+in it::
+
+    C:\Users\user>ipython
+    [...]
+    In [1]: import numpy
+
+If you see an error message telling you that ``DLL load failed``, that is
+probably due to a bug in the script launching ipython. If ``C:\Python27``
+is the directory where you installed EPD, edit
+``C:\Python27\Scripts\ipython.bat``, there should be a line saying::
+
+    set path="C:\Python27";%path%
+
+Remove the quotes around ``Python27``, leading to::
+
+    set path=C:\Python27;%path%
+
+Then, it should work in all new terminals.
+
+pip is not included in EPD, but you can simply install it with::
+
+    easy_install pip
+
+.. _epd: http://enthought.com/products/epd.php
+.. _Graphviz: http://www.graphviz.org/Download_windows.php
+
+Alternative: Python(x,y)
+########################
+
+If you are not eligible to a free download of EPD, and
+neither Python nor MinGW is installed on your computer, the easiest way to
 install most dependencies is to install
 `Python(x,y) <http://www.pythonxy.com>`__.
 It is a single installation
@@ -664,20 +708,28 @@ components as in Python(x,y) that are required by Theano, follow these steps:
 
 At this point, whether you installed Python(x,y) or individual components, you
 should have MinGW, Python, Numpy, Scipy and Nose installed.
-In order to download Theano you will also need Git, which you can get
-`here <http://git-scm.com/download>`__ (download the latest version of the
-"Full installer for official Git" from the ``msysgit`` download page).
+
 
 Installing Theano
 ~~~~~~~~~~~~~~~~~
 
-It should be possible to install Theano with ``pip`` as described in
-:ref:`linux_basic`, but this has not been tested yet under Windows. We describe
+It is possible to install Theano with ``pip`` as described in
+:ref:`linux_basic`. However, if you want to get the latest development version,
+or edit the code, you should follow the instructions below.
+
+Bleeding-edge installation
+##########################
+
+We describe
 here instructions to use the latest code repository version (bleeding-edge).
 Command lines listed below are assumed to be run in a Windows prompt
 (click ``Start`` and type the ``cmd`` command), and may need to be adapted if
 used within a MinGW Shell (not available if you only installed Python(x,y)).
 
+You will need Git, which you can get
+`here <http://git-scm.com/download>`__ (download the latest version of the
+"Full installer for official Git" from the ``msysgit`` download page).
+
 - Navigate into the directory you want Theano to be installed in, and download
   it with
 
@@ -695,6 +747,13 @@ used within a MinGW Shell (not available if you only installed Python(x,y)).
         C:\Users\login>echo %PYTHONPATH%
         C:\Users\login\Theano
 
+Configure Theano
+################
+
+If you installed Python through EPD, you should be all set by now. Otherwise,
+whether you installed Theano through pip or git, you should follow these
+steps.
+
 - Create a new ``.theanorc`` text file (or ``.theanorc.txt``, whichever is easier
   for you to create under Windows) in your user profile directory (the directory you
   are into when you start a new command prompt with ``cmd``), containing the
@@ -762,6 +821,9 @@ the Visual Studio editor.
 Compiling a faster BLAS
 ~~~~~~~~~~~~~~~~~~~~~~~
 
+If you installed Python through EPD, Theano will automatically link with the
+MKL library included in EPD, so you should not need to compile your own BLAS.
+
 .. note::
 
    GotoBLAS is no longer maintained and has been released under the BSD

doc/scripts/docgen.py

@@ -57,8 +57,9 @@ import getopt
 from collections import defaultdict
 
 if __name__ == '__main__':
-
-    throot = "/".join(sys.path[0].split("/")[:-2])
+    # Equivalent of sys.path[0]/../..
+    throot = os.path.abspath(
+        os.path.join(sys.path[0], os.pardir, os.pardir))
 
     options = defaultdict(bool)
     options.update(dict([x, y or True] for x, y in
@@ -91,7 +92,7 @@ if __name__ == '__main__':
 
     # Make sure the appropriate 'theano' directory is in the PYTHONPATH
     pythonpath = os.environ.get('PYTHONPATH', '')
-    pythonpath = throot + ':' + pythonpath
+    pythonpath = os.pathsep.join([throot, pythonpath])
     os.environ['PYTHONPATH'] = pythonpath
 
     if options['--all'] or options['--epydoc']:

theano/compile/tests/test_shared.py

@@ -250,7 +250,7 @@ class Test_SharedVariable(unittest.TestCase):
         f(b, 8)
         assert b.get_value() == 8
 
-        b = shared(numpy.zeros((5, 5), dtype='float32'), allow_downcast=True)
+        b = shared(numpy.zeros((5, 5), dtype='float32'))
         self.assertRaises(TypeError, f, b, numpy.random.rand(5, 5))
 
     def test_tensor_floatX(self):
@@ -303,6 +303,5 @@ class Test_SharedVariable(unittest.TestCase):
         f(b, [8])
         assert b.get_value() == 8
 
-        b = shared(numpy.zeros((5, 5), dtype='float32'),
-                allow_downcast=True)
+        b = shared(numpy.zeros((5, 5), dtype='float32'))
         self.assertRaises(TypeError, f, b, numpy.random.rand(5, 5))

theano/gof/cmodule.py

@@ -1429,6 +1429,13 @@ class GCC_compiler(object):
             # Under Windows it looks like fPIC is useless. Compiler warning:
             # '-fPIC ignored for target (all code is position independent)'
             preargs.append('-fPIC')
+
+        if sys.platform == 'win32' and local_bitwidth() == 64:
+            # Under 64-bits windows installation, sys.platform is 'win32'.
+            # We need to define MS_WIN64 for the preprocessor to be able to
+            # link with libpython.
+            preargs.append('-DMS_WIN64')
+
         no_opt = False
 
         include_dirs = include_dirs + std_include_dirs()

theano/tensor/basic.py

@@ -341,10 +341,13 @@ def constant_or_value(x, rtype, name=None, ndim=None, dtype=None):
         elif rtype is TensorConstant and isinstance(x, float):
             x_ = autocast_float(x)
         elif rtype is TensorConstant and isinstance(x, long):
-            # It is not clear what would happen if one was to use a `long`
-            # number as a constant in a Theano graph. As a result, we throw
-            # an exception in this situation.
-            raise NotImplementedError('Constants of type `long` not supported')
+            # We need to address the case where a long number is used in a
+            # Theano graph, because on Windows 64, all shapes are expressed
+            # with longs.
+            # If a long fits in int64, we convert it into an int64, like
+            # numpy.asarray() does.
+            # If x is too big, an OverflowError will be raised by numpy.
+            x_ = theano._asarray(x, dtype='int64')
         elif isinstance(x, numpy.ndarray):
             x_ = x
             # Currently we do not have a bool dtype in Theano.