提交 · 5c504099843abde0586c9d12762065ee238cdbd1 · testgroup / pytensor

30 1月, 2014 1 次提交
- Remove check_preallocated_output and use default mode for theano.function. · 5c504099
  由 Marc-Alexandre Cote 提交于 1月 29, 2014
  
  5c504099
29 1月, 2014 3 次提交
- Simplify unit tests using check_preallocated_output · 09d7318f
  由 Marc-Alexandre Cote 提交于 1月 28, 2014
  
  09d7318f
- Implemented cumprod function similar to numpy's one. · 195bee05
  由 Marc-Alexandre Cote 提交于 1月 28, 2014
  
  195bee05
- Reuse allocated memory when possible. · 100d336e
  由 Marc-Alexandre Cote 提交于 1月 28, 2014
  
  100d336e
26 1月, 2014 2 次提交
- Changed PyArray_SHAPE to PyArray_DIMS. · 2dddf4ac
  由 Marc-Alexandre Cote 提交于 1月 25, 2014
  
  2dddf4ac
- Added a C version for the CumsumOp. · c9d6cfe0
  由 Marc-Alexandre Cote 提交于 1月 25, 2014
  
  c9d6cfe0
19 1月, 2014 1 次提交
- Removed forgotten debug comments. · 6a99f986
  由 Marc-Alexandre Cote 提交于 1月 18, 2014
  
  6a99f986
18 1月, 2014 1 次提交
- Added the cumsum function similar to numpy's one. · 6850329e
  由 Marc-Alexandre Cote 提交于 1月 18, 2014
  
  6850329e
16 1月, 2014 1 次提交
- Merge pull request #1692 from abergeron/mac_link · e8f6cb73
  由 Frédéric Bastien 提交于 1月 15, 2014
```
Fix frequent segfaults while using the gpu on Mac.
```
  e8f6cb73
14 1月, 2014 4 次提交
- Merge pull request #1696 from lamblin/fix_scan_valueerror · e7b90016
  由 Razvan Pascanu 提交于 1月 13, 2014
```
Do not use arrays as conditions, explicitly test against None
```
  e7b90016
- Do not use arrays as conditions, explicitly test against None · 65c3a49b
  由 Pascal Lamblin 提交于 1月 13, 2014
  
  65c3a49b
- Merge pull request #1695 from abergeron/fix_npcompat · 31ddaabe
  由 Frédéric Bastien 提交于 1月 13, 2014
```
Fix include path for numpy_compat.h following change in compyte.
```
  31ddaabe
- Fix include path for numpy_compat.h following change in compyte. · 957929a3
  由 Arnaud Bergeron 提交于 1月 13, 2014
  
  957929a3
11 1月, 2014 3 次提交
- Merge pull request #1694 from nouiz/prod_no_zeros_in_inputs · fd69ca93
  由 Pascal Lamblin 提交于 1月 10, 2014
```
Make the no_zeros_in_inputs parameter of Prod() available to the user.
```
  fd69ca93
- Fix display of the doc of prod() · 1fb8960b
  由 Frederic 提交于 1月 10, 2014
  
  1fb8960b
- Make the no_zeros_in_inputs parameter of Prod() available to the user. · 218e03c5
  由 Frederic 提交于 1月 10, 2014
  
  218e03c5
10 1月, 2014 2 次提交
- Remove the test file for nvcc_compiler since the only tested function was the one that was removed. · 643bd310
  由 Arnaud Bergeron 提交于 1月 09, 2014
  
  643bd310
- Fix frequent segfaults while using the gpu on Mac. · 8e984c50
  由 Arnaud Bergeron 提交于 1月 09, 2014
  
  8e984c50
09 1月, 2014 2 次提交
- Merge pull request #1691 from SuperElectric/sparse_stack_fix · 4a12110c
  由 Frédéric Bastien 提交于 1月 08, 2014
```
Bugfix in sparse.hstack(), sparse.vstack()
```
  4a12110c
- Merge pull request #1668 from nouiz/gpu_conv · 1d9767b5
  由 abergeron 提交于 1月 08, 2014
```
port GpuConv to the new gpu back-end
```
  1d9767b5
08 1月, 2014 17 次提交
- added unit tests for theano.sparse.hstack(), vstack() · 4b1ece1b
  由 Matthew Koichi Grimes 提交于 1月 08, 2014
  
  4b1ece1b
- Use the good type. · 999f6399
  由 Frederic 提交于 1月 08, 2014
  
  999f6399
- Fix crash in the new gpuconv opt. Make it reuse op_lifter · 5682d3f8
  由 Frederic 提交于 1月 08, 2014
  
  5682d3f8
- pep8 · 024df028
  由 Frederic 提交于 1月 08, 2014
  
  024df028
- Merge branch 'master' into sparse_stack_fix · e2bd5b94
  由 Matthew Koichi Grimes 提交于 1月 08, 2014
  
  e2bd5b94
- remove deprecated comment · a1ddd267
  由 Frederic 提交于 1月 07, 2014
  
  a1ddd267
- implement GpuArrayType.values_eq_approx() parameter atol and rtol. · 8320763f
  由 Frederic 提交于 1月 07, 2014
```
Make the gpuarray opt for conv use it.
```
  8320763f
- pep8 · 66b27afb
  由 Frederic 提交于 1月 07, 2014
  
  66b27afb
- Fix GpuArrayType.values_eq_approx(). It wasn't checking the signe correctly. · f058a81e
  由 Frederic 提交于 1月 07, 2014
  
  f058a81e
- Clean up types in printfs to make sure the printouts will work on 32 and 64 bits machines. · f38a3f3d
  由 Arnaud Bergeron 提交于 1月 07, 2014
  
  f38a3f3d
- First pass at type safety in printfs. · d19a4777
  由 Arnaud Bergeron 提交于 1月 07, 2014
  
  d19a4777
- make sure the input var are of the right type. · a35f07e3
  由 Frederic 提交于 12月 13, 2013
  
  a35f07e3
- new GpuConv compile, but give wrong version in some cases! · 5fc89c03
  由 Frederic 提交于 12月 13, 2013
  
  5fc89c03
- Advance the new GpuConv · baf12f54
  由 Frederic 提交于 12月 13, 2013
  
  baf12f54
- Copy the old GpuConv to the new back-end. · 50677701
  由 Frederic 提交于 12月 13, 2013
  
  50677701
- Merge pull request #1683 from nouiz/opfromgraph · defa003a
  由 abergeron 提交于 1月 07, 2014
```
OpFromGraph documentation and cleanup
```
  defa003a
- Merge pull request #1664 from nouiz/gpu_inc_sub · e05e801f
  由 abergeron 提交于 1月 07, 2014
```
GpuIncSubtensor
```
  e05e801f
07 1月, 2014 3 次提交
- fixed sparse.hstack, sparse.vstack to call scalar.upcast(*args), rather than scalar.upcast(args) · 39eedf7d
  由 Matthew Koichi Grimes 提交于 1月 07, 2014
  
  39eedf7d
- Merge pull request #1689 from nouiz/fix_test · ff27c77a
  由 Pascal Lamblin 提交于 1月 06, 2014
```
Fix test in the daily buildbot by doing the comparison on ndarray, not C...
```
  ff27c77a
- Fix test in the daily buildbot by doing the comparison on ndarray, not CudaNdarray. · 69559928
  由 Frederic 提交于 1月 06, 2014
  
  69559928