CCPi-Regularisation-Toolkit

CCPi-Regularisation-Toolkit copied to clipboard

Implements a faster TNV, by using glib2 and by setting some more aggressive compiler flags. Also uses some memory alignment. It's been tested on a large dataset from IMAT and...

paskino

TNV regulariser does not have any unit test.

paskino

4d extension

2

As @epapoutsellis mentioned, would the 4d extension be useful to incorporate into the toolkit? The obvious obstacle is the memory limits, so X-ray data might not be realistic to process....

dkazanc

moved files in include and install include

1

closes #114 Still needs to be tested

paskino

How to handle CUDA dependency

3

So far the only package depending on CUDA is the regularisation. Currently we rely on the host to have CUDA installed. Jenkins doesn’t build for CUDA, though I seem to...

paskino

Currently the includes are not installed so that the shared library cannot be used by external projects. Something like [this](https://gitlab.com/drFaustroll/cmake-tutorial/blob/master/code/02.3-dot/cxx/src/CMakeLists.txt) should be implemented. ```cmake install(FILES ${PROJECT_SOURCE_DIR}/include/algebra.h DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}) ``` Additionally,...

paskino

C functions should return int or void

3

`int` would be better so that basic checks could be implemented

paskino

Spatially variant regularisation field

10

* One requires a capability of passing not only a scalar but a vector of regularisation values. * When the N-dimensional field of regularisation parameters is evaluated, it can be...

dkazanc

Create virtual algorithm class and concrete algorithm classes

3

Currently the algorithms used in the toolkit are repeated within the different functions. This code repetition may lead to error. Let's create an iterative algorithm class with the same structure...

paskino

After cmake build and running GPU modules in Matlab (Linux). The related [issue](https://github.com/ContinuumIO/anaconda-issues/issues/5191). Can be fixed either linking to a newer `libstdc++.so.6.0.24` or by setting the env variable: `export LD_PRELOAD=/home/user/anaconda3/lib/libstdc++.so.6.0.24`...

dkazanc