Pierre Yger
Pierre Yger
This PR depends on Sam's one, with the peak pipeline. I need to finish some tests, but this is running !
The sparse matrix is now an external module, that can be loaded separatly such that cudamat does not depend on scipy. There is also a test for sparse_dot product.
Sounds like the perfect package, handling sparse matrices
It would be nice to be able to launch the code on a list of filenames, provided by the users, without the need to have a .params next to a...
Currently, the software is able to work with 3D probes, but none of the GUIs (preview, Matlab, phy) will display the data.
There is a problem with MPI 3.0 and the code, because of recursive calls to mpiexec. Need to be fixed or to pin version
We should refactor the code, when neo.rawio will be out, rely on neo as a dependence and have a common way to read/write proprietary file formats. This would make everthing...
Currently, since we made the choice of sparse matrices everywhere, the GPU usage is not really efficient and do not bring much compared to several CPUs. However, by re-writing custom...