{ "id": 4501, "title": "Running python from R", "url": "/competitions/challenges-in-representation-learning-the-black-box-learning-challenge/discussion/4501", "author_name": "", "post_date": "2013-05-04T12:08:43.660Z", "votes": null, "comment_count": 5, "views": 6033, "content": "

I'm trying run the python scripts from R. Something like this:

\r\n
setwd('c:/anaconda/lib/site-packages/pylearn2/scripts/icml_2013_wrepl/black_box')
system('cd c:/anaconda/lib/site-packages/pylearn2/scripts/icml_2013_wrepl/black_box')
system('python train.py mlp.yaml')
\r\n

but the compiler fail.

\r\n

(From cmd on windows the python scrips run fine)

\r\n

Some tip for fix it?

\r\n
Problem occurred during compilation with the command line below:
g++ -shared -g -O3 -Wno-write-strings -Wno-unused-label -Wno-unused-variable -fno-math-errno -DMS_WIN64 -m64 -IC:\\Anaconda\\lib\\site-packages\\numpy\\core\\include -IC:\\Anaconda\\include -o C:\\Users\\JAG\\AppData\\Local\\Theano\\compiledir_Windows-7-6.1.7601-SP1-Intel64_Family_6_Model_42_Stepping_7_GenuineIntel-2.7.3\\tmph_sqib\\baf8a6fdee34b135ea9110785e356489.pyd C:\\Users\\JAG\\AppData\\Local\\Theano\\compiledir_Windows-7-6.1.7601-SP1-Intel64_Family_6_Model_42_Stepping_7_GenuineIntel-2.7.3\\tmph_sqib\\mod.cpp -LC:\\Anaconda\\libs -LC:\\Anaconda -lpython27 -lblas
Traceback (most recent call last):
  File "train.py", line 141, in <module>
    train_obj.main_loop()
  File "C:\\Anaconda\\lib\\site-packages\\pylearn2\\train.py", line 109, in main_loop
    self.algorithm.setup(model=self.model, dataset=self.dataset)
  File "C:\\Anaconda\\lib\\site-packages\\pylearn2\\training_algorithms\\sgd.py", line 285, in setup
    mode=self.theano_function_mode)
  File "C:\\Anaconda\\lib\\site-packages\\theano\\compile\\function.py", line 221, in function
    profile=profile)
  File "C:\\Anaconda\\lib\\site-packages\\theano\\compile\\pfunc.py", line 506, in pfunc
    on_unused_input=on_unused_input)
  File "C:\\Anaconda\\lib\\site-packages\\theano\\compile\\function_module.py", line 1339, in orig_function
    defaults)
  File "C:\\Anaconda\\lib\\site-packages\\theano\\compile\\function_module.py", line 1167, in create
    _fn, _i, _o = self.linker.make_thunk(input_storage=input_storage_lists)
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\link.py", line 383, in make_thunk
    output_storage = output_storage)[:3]
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\vm.py", line 799, in make_all
    for node in order]
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\op.py", line 580, in make_thunk
    output_storage=node_output_storage)
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\cc.py", line 913, in make_thunk
    keep_lock=keep_lock)
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\cc.py", line 856, in __compile__
    keep_lock=keep_lock)
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\cc.py", line 1279, in cthunk_factory
    key=key, fn=self.compile_cmodule_by_step, keep_lock=keep_lock)
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\cmodule.py", line 975, in module_from_key
    module = compile_steps.next()
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\cc.py", line 1201, in compile_cmodule_by_step
    preargs=preargs)
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\cmodule.py", line 1633, in compile_str
    (status, compile_stderr.replace('\\n', '. ')))
Exception: ('Compilation failed (return status=1): In file included from C:\\\\Anaconda\\\\lib\\\\site-packages\\\\numpy\\\\core\\\\include/numpy/ndarraytypes.h:1728:0,\\r.                  from C:\\\\Anaconda\\\\lib\\\\site-packages\\\\numpy\\\\core\\\\include/numpy/ndarrayobject.h:17,\\r.                  from C:\\\\Anaconda\\\\lib\\\\site-packages\\\\numpy\\\\core\\\\include/numpy/arrayobject.h:15,\\r.                  from C:\\\\Users\\\\JAG\\\\AppData\\\\Local\\\\Theano\\\\compiledir_Windows-7-6.1.7601-SP1-Intel64_Family_6_Model_42_Stepping_7_GenuineIntel-2.7.3\\\\tmph_sqib\\\\mod.cpp:3:\\r. C:\\\\Anaconda\\\\lib\\\\site-packages\\\\numpy\\\\core\\\\include/numpy/npy_deprecated_api.h:9:82: note: #pragma message: C:\\\\Anaconda\\\\lib\\\\site-packages\\\\numpy\\\\core\\\\include/numpy/npy_deprecated_api.h(8) : Warning Msg: Using deprecated NumPy API, disable it by #defining NPY_NO_DEPRECATED_API NPY_1_7_API_VERSION\\r. c:/anaconda/mingw/bin/../lib/gcc/x86_64-w64-mingw32/4.7.0/../../../../x86_64-w64-mingw32/bin/ld.exe: cannot find -lblas\\r. collect2.exe: error: ld returned 1 exit status\\r. ', '[_dot22(SGD[X], h0_W)]')
Mensajes de aviso perdidos
comando ejecutado 'python train.py mlp.yaml' tiene estatus 1
\r\n
 
", "messages": [ { "id": "23889", "postDate": "05/04/2013 12:08:43", "content": "

I'm trying run the python scripts from R. Something like this:

\r\n
setwd('c:/anaconda/lib/site-packages/pylearn2/scripts/icml_2013_wrepl/black_box')
system('cd c:/anaconda/lib/site-packages/pylearn2/scripts/icml_2013_wrepl/black_box')
system('python train.py mlp.yaml')
\r\n

but the compiler fail.

\r\n

(From cmd on windows the python scrips run fine)

\r\n

Some tip for fix it?

\r\n
Problem occurred during compilation with the command line below:
g++ -shared -g -O3 -Wno-write-strings -Wno-unused-label -Wno-unused-variable -fno-math-errno -DMS_WIN64 -m64 -IC:\\Anaconda\\lib\\site-packages\\numpy\\core\\include -IC:\\Anaconda\\include -o C:\\Users\\JAG\\AppData\\Local\\Theano\\compiledir_Windows-7-6.1.7601-SP1-Intel64_Family_6_Model_42_Stepping_7_GenuineIntel-2.7.3\\tmph_sqib\\baf8a6fdee34b135ea9110785e356489.pyd C:\\Users\\JAG\\AppData\\Local\\Theano\\compiledir_Windows-7-6.1.7601-SP1-Intel64_Family_6_Model_42_Stepping_7_GenuineIntel-2.7.3\\tmph_sqib\\mod.cpp -LC:\\Anaconda\\libs -LC:\\Anaconda -lpython27 -lblas
Traceback (most recent call last):
  File "train.py", line 141, in <module>
    train_obj.main_loop()
  File "C:\\Anaconda\\lib\\site-packages\\pylearn2\\train.py", line 109, in main_loop
    self.algorithm.setup(model=self.model, dataset=self.dataset)
  File "C:\\Anaconda\\lib\\site-packages\\pylearn2\\training_algorithms\\sgd.py", line 285, in setup
    mode=self.theano_function_mode)
  File "C:\\Anaconda\\lib\\site-packages\\theano\\compile\\function.py", line 221, in function
    profile=profile)
  File "C:\\Anaconda\\lib\\site-packages\\theano\\compile\\pfunc.py", line 506, in pfunc
    on_unused_input=on_unused_input)
  File "C:\\Anaconda\\lib\\site-packages\\theano\\compile\\function_module.py", line 1339, in orig_function
    defaults)
  File "C:\\Anaconda\\lib\\site-packages\\theano\\compile\\function_module.py", line 1167, in create
    _fn, _i, _o = self.linker.make_thunk(input_storage=input_storage_lists)
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\link.py", line 383, in make_thunk
    output_storage = output_storage)[:3]
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\vm.py", line 799, in make_all
    for node in order]
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\op.py", line 580, in make_thunk
    output_storage=node_output_storage)
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\cc.py", line 913, in make_thunk
    keep_lock=keep_lock)
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\cc.py", line 856, in __compile__
    keep_lock=keep_lock)
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\cc.py", line 1279, in cthunk_factory
    key=key, fn=self.compile_cmodule_by_step, keep_lock=keep_lock)
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\cmodule.py", line 975, in module_from_key
    module = compile_steps.next()
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\cc.py", line 1201, in compile_cmodule_by_step
    preargs=preargs)
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\cmodule.py", line 1633, in compile_str
    (status, compile_stderr.replace('\\n', '. ')))
Exception: ('Compilation failed (return status=1): In file included from C:\\\\Anaconda\\\\lib\\\\site-packages\\\\numpy\\\\core\\\\include/numpy/ndarraytypes.h:1728:0,\\r.                  from C:\\\\Anaconda\\\\lib\\\\site-packages\\\\numpy\\\\core\\\\include/numpy/ndarrayobject.h:17,\\r.                  from C:\\\\Anaconda\\\\lib\\\\site-packages\\\\numpy\\\\core\\\\include/numpy/arrayobject.h:15,\\r.                  from C:\\\\Users\\\\JAG\\\\AppData\\\\Local\\\\Theano\\\\compiledir_Windows-7-6.1.7601-SP1-Intel64_Family_6_Model_42_Stepping_7_GenuineIntel-2.7.3\\\\tmph_sqib\\\\mod.cpp:3:\\r. C:\\\\Anaconda\\\\lib\\\\site-packages\\\\numpy\\\\core\\\\include/numpy/npy_deprecated_api.h:9:82: note: #pragma message: C:\\\\Anaconda\\\\lib\\\\site-packages\\\\numpy\\\\core\\\\include/numpy/npy_deprecated_api.h(8) : Warning Msg: Using deprecated NumPy API, disable it by #defining NPY_NO_DEPRECATED_API NPY_1_7_API_VERSION\\r. c:/anaconda/mingw/bin/../lib/gcc/x86_64-w64-mingw32/4.7.0/../../../../x86_64-w64-mingw32/bin/ld.exe: cannot find -lblas\\r. collect2.exe: error: ld returned 1 exit status\\r. ', '[_dot22(SGD[X], h0_W)]')
Mensajes de aviso perdidos
comando ejecutado 'python train.py mlp.yaml' tiene estatus 1
\r\n
 
", "rawMarkdown": "", "votes": null }, { "id": "23955", "postDate": "05/06/2013 21:45:19", "content": "

I am having a problem similar to Jose.  I am running into problems running the benchmark pylearn2 python scripts from within R, but they work fine from the Windows command line.  Any chance someone knows how to solve this?

\r\n

I am running Anaconda python on Windows 7 64.

\r\n

 

\r\n

The problem occurs right as it reports "Compiling sgd_update..."

\r\n

 

\r\n
\r\n
 
\r\n
> shell("C:/Archive/AWS/BlackBox/python_batch_R.bat", intern=FALSE, translate=TRUE)
\r\n
 
\r\n
C:\\Anaconda\\Lib\\site-packages\\pylearn2\\scripts\\icml_2013_wrepl\\black_box>cd c:\\Anaconda\\Lib\\site-packages\\pylearn2\\scripts\\icml_2013_wrepl\\black_box
\r\n
 
\r\n
C:\\Anaconda\\Lib\\site-packages\\pylearn2\\scripts\\icml_2013_wrepl\\black_box>python ..\\..\\train.py mlp_R.yaml
\r\n
Parameter and initial learning rate summary:
\r\n
 
\r\n
C:\\Anaconda\\Lib\\site-packages\\pylearn2\\scripts\\icml_2013_wrepl\\black_box>cd c:\\Anaconda\\Lib\\site-packages\\pylearn2\\scripts\\icml_2013_wrepl\\black_box
\r\n
 
\r\n
C:\\Anaconda\\Lib\\site-packages\\pylearn2\\scripts\\icml_2013_wrepl\\black_box>python ..\\..\\train.py mlp_R.yaml
\r\n
Parameter and initial learning rate summary:
\r\n
C:\\Anaconda\\lib\\site-packages\\pylearn2\\models\\mlp.py:40: UserWarning: MLP changing the recursion limit.
\r\n
  warnings.warn("MLP changing the recursion limit.")
\r\n
        h0_W: 0.1
\r\n
        h0_b: 0.1
\r\n
        h1_W: 0.1
\r\n
        h1_b: 0.1
\r\n
        softmax_b: 0.1
\r\n
        softmax_W: 0.1
\r\n
Compiling sgd_update...
\r\n
00001   #include <Python.h>
\r\n
00002   #include <iostream>
\r\n
00003   #include <numpy/arrayobject.h>
\r\n
00004   #include <iostream>
\r\n
00005   #include <math.h>
\r\n
00006   #include <sys/time.h>
\r\n
00007   #include <numpy/arrayscalars.h>
\r\n
00008   #include <time.h>
\r\n
00009   //////////////////////
\r\n
00010   ////  Support Code
\r\n
00011   //////////////////////
\r\n
00012  
\r\n
00013  
\r\n
00014       extern "C"
\r\n

 

\r\n

….

\r\n

 

\r\n
0845   
\r\n
00846   //////////////////////
\r\n
00847   ////  Module init
\r\n
00848   //////////////////////
\r\n
00849   static PyMethodDef MyMethods[] = {
\r\n
00850          {"instantiate", instantiate, METH_VARARGS, "undocumented"} ,
\r\n
00851          {NULL, NULL, 0, NULL}
\r\n
00852   };
\r\n
00853   PyMODINIT_FUNC initbaf8a6fdee34b135ea9110785e356489(void){
\r\n
00854      import_array();
\r\n
00855      (void) Py_InitModule("baf8a6fdee34b135ea9110785e356489", MyMethods);
\r\n
00856   }
\r\n
00857  
\r\n
Problem occurred during compilation with the command line below:
\r\n
g++ -shared -g -O3 -Wno-write-strings -Wno-unused-label -Wno-unused-variable -fno-math-errno -DMS_WIN64 -m64 -IC:\\Anaconda\\lib\\site-packages\\numpy\\core\\include -IC:\\Anaconda\\include -o C:\\Users\\Nathan\\AppData\\Local\\Theano\\compiledir_Windows-7-6.1.7601-SP1-Intel64_Family_6_Model_45_Stepping_7_GenuineIntel-2.7.3\\tmpje9pin\\baf8a6fdee34b135ea9110785e356489.pyd C:\\Users\\Nathan\\AppData\\Local\\Theano\\compiledir_Windows-7-6.1.7601-SP1-Intel64_Family_6_Model_45_Stepping_7_GenuineIntel-2.7.3\\tmpje9pin\\mod.cpp -LC:\\Anaconda\\libs -LC:\\Anaconda -lpython27 -lblas
\r\n
Traceback (most recent call last):
\r\n
  File "..\\..\\train.py", line 141, in <module>
\r\n
    train_obj.main_loop()
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\pylearn2\\train.py", line 109, in main_loop
\r\n
    self.algorithm.setup(model=self.model, dataset=self.dataset)
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\pylearn2\\training_algorithms\\sgd.py", line 285, in setup
\r\n
    mode=self.theano_function_mode)
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\theano\\compile\\function.py", line 221, in function
\r\n
    profile=profile)
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\theano\\compile\\pfunc.py", line 506, in pfunc
\r\n
    on_unused_input=on_unused_input)
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\theano\\compile\\function_module.py", line 1339, in orig_function
\r\n
    defaults)
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\theano\\compile\\function_module.py", line 1167, in create
\r\n
    _fn, _i, _o = self.linker.make_thunk(input_storage=input_storage_lists)
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\link.py", line 383, in make_thunk
\r\n
    output_storage = output_storage)[:3]
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\vm.py", line 799, in make_all
\r\n
    for node in order]
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\op.py", line 580, in make_thunk
\r\n
    output_storage=node_output_storage)
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\cc.py", line 913, in make_thunk
\r\n
    keep_lock=keep_lock)
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\cc.py", line 856, in __compile__
\r\n
    keep_lock=keep_lock)
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\cc.py", line 1279, in cthunk_factory
\r\n
    key=key, fn=self.compile_cmodule_by_step, keep_lock=keep_lock)
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\cmodule.py", line 975, in module_from_key
\r\n
    module = compile_steps.next()
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\cc.py", line 1201, in compile_cmodule_by_step
\r\n
    preargs=preargs)
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\cmodule.py", line 1633, in compile_str
\r\n
    (status, compile_stderr.replace('\\n', '. ')))
\r\n
Exception: ('Compilation failed (return status=1): In file included from C:\\\\Anaconda\\\\lib\\\\site-packages\\\\numpy\\\\core\\\\include/numpy/ndarraytypes.h:1728:0,\\r.                  from C:\\\\Anaconda\\\\lib\\\\site-packages\\\\numpy\\\\core\\\\include/numpy/ndarrayobject.h:17,\\r.                  from C:\\\\Anaconda\\\\lib\\\\site-packages\\\\numpy\\\\core\\\\include/numpy/arrayobject.h:15,\\r.                  from C:\\\\Users\\\\Nathan\\\\AppData\\\\Local\\\\Theano\\\\compiledir_Windows-7-6.1.7601-SP1-Intel64_Family_6_Model_45_Stepping_7_GenuineIntel-2.7.3\\\\tmpje9pin\\\\mod.cpp:3:\\r. C:\\\\Anaconda\\\\lib\\\\site-packages\\\\numpy\\\\core\\\\include/numpy/npy_deprecated_api.h:9:82: note: #pragma message: C:\\\\Anaconda\\\\lib\\\\site-packages\\\\numpy\\\\core\\\\include/numpy/npy_deprecated_api.h(8) : Warning Msg: Using deprecated NumPy API, disable it by #defining NPY_NO_DEPRECATED_API NPY_1_7_API_VERSION\\r. c:/anaconda/mingw/bin/../lib/gcc/x86_64-w64-mingw32/4.7.0/../../../../x86_64-w64-mingw32/bin/ld.exe: cannot find -lblas\\r. collect2.exe: error: ld returned 1 exit status\\r. ', '[_dot22(SGD[X], h0_W)]')
\r\n

 

\r\n


\r\n

\r\n
\r\n

 

", "rawMarkdown": "", "votes": null }, { "id": "23962", "postDate": "05/06/2013 22:56:43", "content": "

I don't know how to do this. Maybe someone else will. As a pylearn2 developer, I'm curious why you want to be able to run pylearn2 from inside R though. Is it just for the convenience of not switching shells? Or can R use the console output somehow? I imagine\r\n R can't use the pickle file it emits.

", "rawMarkdown": "", "votes": null }, { "id": "23963", "postDate": "05/06/2013 22:57:11", "content": "

Hi Jose,

\r\n

    Have you tried creating a .theanorc.txt file with a "blas=" line in it?

", "rawMarkdown": "", "votes": null }, { "id": "23990", "postDate": "05/07/2013 18:09:23", "content": "

[quote=Ian Goodfellow;23962]

\r\n

I don't know how to do this. Maybe someone else will. As a pylearn2 developer, I'm curious why you want to be able to run pylearn2 from inside R though. Is it just for the convenience of not switching shells? Or can R use the console output somehow? I imagine\r\n R can't use the pickle file it emits.

\r\n

[/quote]

\r\n

Ian, only for data manage in R. I'm learning python with this challenge... python,numpy,theano,pylearn2. I spend a lot of time to do things  in python that I know how to do in R in seconds. This was the reason.

\r\n

Daniel: Thank you,  I decided do all in python... (?!%&!rrrr).

", "rawMarkdown": "", "votes": null }, { "id": "23998", "postDate": "05/07/2013 21:44:34", "content": "

Problem is that .theanorc env is not visible to theano
\r\n
\r\neasy, but not best workaround is:
\r\n
\r\n1. create  StartFromR.bat
\r\nwith:
\r\n    train.py test7.yaml
\r\n2. in train.py set variables (variables from .theanorc)
\r\n     edit train.py:

\r\n
from pylearn2.utils.timing import log_timing

import theano
theano.config.blas.cxxflags = '-IC:\\mingw64\\include'
theano.config.blas.ldflags = ''

log = logging.getLogger(__name__)
\r\n

3. run in R
\r\n> shell('StartFromR.bat')
\r\n
\r\n....

\r\n
Saving to mlp2007.pkl...
Saving to mlp2007.pkl done. Time elapsed: 1.833000 seconds
Saving to mlp2007.pkl...
Saving to mlp2007.pkl done. Time elapsed: 1.803000 seconds
>
\r\n


\r\n      

", "rawMarkdown": "", "votes": null } ], "comments": [ { "id": 23955, "author_name": "bignate", "author_url": "", "post_date": "05/06/2013 21:45:19", "content": "

I am having a problem similar to Jose.  I am running into problems running the benchmark pylearn2 python scripts from within R, but they work fine from the Windows command line.  Any chance someone knows how to solve this?

\r\n

I am running Anaconda python on Windows 7 64.

\r\n

 

\r\n

The problem occurs right as it reports "Compiling sgd_update..."

\r\n

 

\r\n
\r\n
 
\r\n
> shell("C:/Archive/AWS/BlackBox/python_batch_R.bat", intern=FALSE, translate=TRUE)
\r\n
 
\r\n
C:\\Anaconda\\Lib\\site-packages\\pylearn2\\scripts\\icml_2013_wrepl\\black_box>cd c:\\Anaconda\\Lib\\site-packages\\pylearn2\\scripts\\icml_2013_wrepl\\black_box
\r\n
 
\r\n
C:\\Anaconda\\Lib\\site-packages\\pylearn2\\scripts\\icml_2013_wrepl\\black_box>python ..\\..\\train.py mlp_R.yaml
\r\n
Parameter and initial learning rate summary:
\r\n
 
\r\n
C:\\Anaconda\\Lib\\site-packages\\pylearn2\\scripts\\icml_2013_wrepl\\black_box>cd c:\\Anaconda\\Lib\\site-packages\\pylearn2\\scripts\\icml_2013_wrepl\\black_box
\r\n
 
\r\n
C:\\Anaconda\\Lib\\site-packages\\pylearn2\\scripts\\icml_2013_wrepl\\black_box>python ..\\..\\train.py mlp_R.yaml
\r\n
Parameter and initial learning rate summary:
\r\n
C:\\Anaconda\\lib\\site-packages\\pylearn2\\models\\mlp.py:40: UserWarning: MLP changing the recursion limit.
\r\n
  warnings.warn("MLP changing the recursion limit.")
\r\n
        h0_W: 0.1
\r\n
        h0_b: 0.1
\r\n
        h1_W: 0.1
\r\n
        h1_b: 0.1
\r\n
        softmax_b: 0.1
\r\n
        softmax_W: 0.1
\r\n
Compiling sgd_update...
\r\n
00001   #include <Python.h>
\r\n
00002   #include <iostream>
\r\n
00003   #include <numpy/arrayobject.h>
\r\n
00004   #include <iostream>
\r\n
00005   #include <math.h>
\r\n
00006   #include <sys/time.h>
\r\n
00007   #include <numpy/arrayscalars.h>
\r\n
00008   #include <time.h>
\r\n
00009   //////////////////////
\r\n
00010   ////  Support Code
\r\n
00011   //////////////////////
\r\n
00012  
\r\n
00013  
\r\n
00014       extern "C"
\r\n

 

\r\n

….

\r\n

 

\r\n
0845   
\r\n
00846   //////////////////////
\r\n
00847   ////  Module init
\r\n
00848   //////////////////////
\r\n
00849   static PyMethodDef MyMethods[] = {
\r\n
00850          {"instantiate", instantiate, METH_VARARGS, "undocumented"} ,
\r\n
00851          {NULL, NULL, 0, NULL}
\r\n
00852   };
\r\n
00853   PyMODINIT_FUNC initbaf8a6fdee34b135ea9110785e356489(void){
\r\n
00854      import_array();
\r\n
00855      (void) Py_InitModule("baf8a6fdee34b135ea9110785e356489", MyMethods);
\r\n
00856   }
\r\n
00857  
\r\n
Problem occurred during compilation with the command line below:
\r\n
g++ -shared -g -O3 -Wno-write-strings -Wno-unused-label -Wno-unused-variable -fno-math-errno -DMS_WIN64 -m64 -IC:\\Anaconda\\lib\\site-packages\\numpy\\core\\include -IC:\\Anaconda\\include -o C:\\Users\\Nathan\\AppData\\Local\\Theano\\compiledir_Windows-7-6.1.7601-SP1-Intel64_Family_6_Model_45_Stepping_7_GenuineIntel-2.7.3\\tmpje9pin\\baf8a6fdee34b135ea9110785e356489.pyd C:\\Users\\Nathan\\AppData\\Local\\Theano\\compiledir_Windows-7-6.1.7601-SP1-Intel64_Family_6_Model_45_Stepping_7_GenuineIntel-2.7.3\\tmpje9pin\\mod.cpp -LC:\\Anaconda\\libs -LC:\\Anaconda -lpython27 -lblas
\r\n
Traceback (most recent call last):
\r\n
  File "..\\..\\train.py", line 141, in <module>
\r\n
    train_obj.main_loop()
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\pylearn2\\train.py", line 109, in main_loop
\r\n
    self.algorithm.setup(model=self.model, dataset=self.dataset)
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\pylearn2\\training_algorithms\\sgd.py", line 285, in setup
\r\n
    mode=self.theano_function_mode)
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\theano\\compile\\function.py", line 221, in function
\r\n
    profile=profile)
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\theano\\compile\\pfunc.py", line 506, in pfunc
\r\n
    on_unused_input=on_unused_input)
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\theano\\compile\\function_module.py", line 1339, in orig_function
\r\n
    defaults)
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\theano\\compile\\function_module.py", line 1167, in create
\r\n
    _fn, _i, _o = self.linker.make_thunk(input_storage=input_storage_lists)
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\link.py", line 383, in make_thunk
\r\n
    output_storage = output_storage)[:3]
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\vm.py", line 799, in make_all
\r\n
    for node in order]
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\op.py", line 580, in make_thunk
\r\n
    output_storage=node_output_storage)
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\cc.py", line 913, in make_thunk
\r\n
    keep_lock=keep_lock)
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\cc.py", line 856, in __compile__
\r\n
    keep_lock=keep_lock)
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\cc.py", line 1279, in cthunk_factory
\r\n
    key=key, fn=self.compile_cmodule_by_step, keep_lock=keep_lock)
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\cmodule.py", line 975, in module_from_key
\r\n
    module = compile_steps.next()
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\cc.py", line 1201, in compile_cmodule_by_step
\r\n
    preargs=preargs)
\r\n
  File "C:\\Anaconda\\lib\\site-packages\\theano\\gof\\cmodule.py", line 1633, in compile_str
\r\n
    (status, compile_stderr.replace('\\n', '. ')))
\r\n
Exception: ('Compilation failed (return status=1): In file included from C:\\\\Anaconda\\\\lib\\\\site-packages\\\\numpy\\\\core\\\\include/numpy/ndarraytypes.h:1728:0,\\r.                  from C:\\\\Anaconda\\\\lib\\\\site-packages\\\\numpy\\\\core\\\\include/numpy/ndarrayobject.h:17,\\r.                  from C:\\\\Anaconda\\\\lib\\\\site-packages\\\\numpy\\\\core\\\\include/numpy/arrayobject.h:15,\\r.                  from C:\\\\Users\\\\Nathan\\\\AppData\\\\Local\\\\Theano\\\\compiledir_Windows-7-6.1.7601-SP1-Intel64_Family_6_Model_45_Stepping_7_GenuineIntel-2.7.3\\\\tmpje9pin\\\\mod.cpp:3:\\r. C:\\\\Anaconda\\\\lib\\\\site-packages\\\\numpy\\\\core\\\\include/numpy/npy_deprecated_api.h:9:82: note: #pragma message: C:\\\\Anaconda\\\\lib\\\\site-packages\\\\numpy\\\\core\\\\include/numpy/npy_deprecated_api.h(8) : Warning Msg: Using deprecated NumPy API, disable it by #defining NPY_NO_DEPRECATED_API NPY_1_7_API_VERSION\\r. c:/anaconda/mingw/bin/../lib/gcc/x86_64-w64-mingw32/4.7.0/../../../../x86_64-w64-mingw32/bin/ld.exe: cannot find -lblas\\r. collect2.exe: error: ld returned 1 exit status\\r. ', '[_dot22(SGD[X], h0_W)]')
\r\n

 

\r\n


\r\n

\r\n
\r\n

 

", "votes": null, "replies": [] }, { "id": 23962, "author_name": "iangoodfellow", "author_url": "", "post_date": "05/06/2013 22:56:43", "content": "

I don't know how to do this. Maybe someone else will. As a pylearn2 developer, I'm curious why you want to be able to run pylearn2 from inside R though. Is it just for the convenience of not switching shells? Or can R use the console output somehow? I imagine\r\n R can't use the pickle file it emits.

", "votes": null, "replies": [] }, { "id": 23963, "author_name": "dpineo", "author_url": "", "post_date": "05/06/2013 22:57:11", "content": "

Hi Jose,

\r\n

    Have you tried creating a .theanorc.txt file with a "blas=" line in it?

", "votes": null, "replies": [] }, { "id": 23990, "author_name": "blindape", "author_url": "", "post_date": "05/07/2013 18:09:23", "content": "

[quote=Ian Goodfellow;23962]

\r\n

I don't know how to do this. Maybe someone else will. As a pylearn2 developer, I'm curious why you want to be able to run pylearn2 from inside R though. Is it just for the convenience of not switching shells? Or can R use the console output somehow? I imagine\r\n R can't use the pickle file it emits.

\r\n

[/quote]

\r\n

Ian, only for data manage in R. I'm learning python with this challenge... python,numpy,theano,pylearn2. I spend a lot of time to do things  in python that I know how to do in R in seconds. This was the reason.

\r\n

Daniel: Thank you,  I decided do all in python... (?!%&!rrrr).

", "votes": null, "replies": [] }, { "id": 23998, "author_name": "djajetic", "author_url": "", "post_date": "05/07/2013 21:44:34", "content": "

Problem is that .theanorc env is not visible to theano
\r\n
\r\neasy, but not best workaround is:
\r\n
\r\n1. create  StartFromR.bat
\r\nwith:
\r\n    train.py test7.yaml
\r\n2. in train.py set variables (variables from .theanorc)
\r\n     edit train.py:

\r\n
from pylearn2.utils.timing import log_timing

import theano
theano.config.blas.cxxflags = '-IC:\\mingw64\\include'
theano.config.blas.ldflags = ''

log = logging.getLogger(__name__)
\r\n

3. run in R
\r\n> shell('StartFromR.bat')
\r\n
\r\n....

\r\n
Saving to mlp2007.pkl...
Saving to mlp2007.pkl done. Time elapsed: 1.833000 seconds
Saving to mlp2007.pkl...
Saving to mlp2007.pkl done. Time elapsed: 1.803000 seconds
>
\r\n


\r\n      

", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "23889": "", "23955": "", "23962": "", "23963": "", "23990": "", "23998": "" }, "source": "meta" }