Code

Discrete convolution#1230

Code

Open

kohr-h wants to merge 17 commits intoodlgroup:masterodlgroup/odl:masterfrom kohr-h:issue-209__convolutionkohr-h/odl:issue-209__convolutionCopy head branch name to clipboard

+1,884-63

Commits

Commits on Mar 15, 2019

• MAINT: rename ufunc_ops->oplib as place for "advanced" operators

  

  Holger Kohr

  authored and

  kohr-h

  committed
  86dc321

  Copy full SHA for 86dc321

• STY: fix pep8 issue in ufunc_ops example

  

  Holger Kohr

  authored and

  kohr-h

  committed
  68f8b70

  Copy full SHA for 68f8b70

• MAINT: fix bad oplib/__init__.py

  

  Holger Kohr

  authored and

  kohr-h

  committed
  061453e

  Copy full SHA for 061453e

• ENH: add DiscreteConvolution operator

  

  Holger Kohr

  authored and

  kohr-h

  committed
  a1689a3

  Copy full SHA for a1689a3

• TST: add tests for DiscreteConvolution init and 1d and 2d eval

  

  Holger Kohr

  authored and

  kohr-h

  committed
  d499e13

  Copy full SHA for d499e13

• ENH: add auto_weighting decorator for adjoints

  

  Holger Kohr

  authored and

  kohr-h

  committed
  edbfe65

  Copy full SHA for edbfe65

• ENH: add auto_weighting example

  

  Holger Kohr

  authored and

  kohr-h

  committed
  dcc585d

  Copy full SHA for dcc585d

• TST: add tests for auto_weighting decorator

  

  Holger Kohr

  authored and

  kohr-h

  committed
  d810538

  Copy full SHA for d810538

• MAINT: catch and test real vs complex issues

  

  Holger Kohr

  authored and

  kohr-h

  committed
  b849f94

  Copy full SHA for b849f94

• ENH: implement adjoint of DiscreteConvolution

  

  Holger Kohr

  authored and

  kohr-h

  committed
  e37cc53

  Copy full SHA for e37cc53

• TST: add discrete convolution adjoint tests

  

  Holger Kohr

  authored and

  kohr-h

  committed
  0db25e2

  Copy full SHA for 0db25e2

• ENH: use out efficiently for pyfftw and no padding in convolution

  

  Holger Kohr

  authored and

  kohr-h

  committed
  d17c530

  Copy full SHA for d17c530

• TST: be less strict in convolution adjoint tests

  

  Holger Kohr

  authored and

  kohr-h

  committed
  913ffa3

  Copy full SHA for 913ffa3

• ENH: add backport of np.roll for multiple axes

  kohr-h

  committed
  72d0c9a

  Copy full SHA for 72d0c9a

• ENH: add utility function for convolution

  kohr-h

  committed
  3993bae

  Copy full SHA for 3993bae

• ENH: add correlate helper for cross-correlation

  kohr-h

  committed
  ad80169

  Copy full SHA for ad80169

• DOC: add Return section to convolve doc

  kohr-h

  committed
  7c89f67

  Copy full SHA for 7c89f67