Juan Ignacio Polanco
Juan Ignacio Polanco
Certain inverse FFTs of multidimensional arrays run very slowly, and allocate considerably, when run with at least 4 threads. The following snippet reproduces the issue, with `JULIA_NUM_THREADS=4`: ```julia using Base.Threads...
Under my configuration, there is a memory leak when loading a JLD file with multiple arrays. I wrote a [script](https://gist.github.com/jipolanco/3fd2bb462f1381e779f1c8440608f062) which can help test the issue. The issue is observed...
Things to do: - [x] define `PencilFFTPlan` constructor taking an existing `PencilArray`; - [ ] define planner functions, such as `plan_fft`, as wrappers of `PencilFFTPlan` constructor. Also extend r2r plans...
Not that simple due to the change from a real to a complex element type. This will require changes to `ManyPencilArray` in PencilArrays, to allow for variable element types of...
For out-of-place FFTs, it should be possible to perform some of the intermediate FFTs in-place, which may lead to memory savings and maybe a small speed-up. For instance, a 3D...
See for instance the [FFTW docs](http://www.fftw.org/pruned.html). This is supported by P3DFFT2 and PFFT.
First of all, thank you for this amazing library and for providing a Julia interface. This PR is to propose a few minor changes with the aim of suppressing some...
To do: - [x] add and fix tests - [ ] document tensor-product splines - [x] add multidimensional interpolations - [ ] support derivatives and integrals?
Add support for exchanging neighbouring data between MPI processes. - The user should be able to select the number of planes to exchange along each decomposed dimension. - In principle,...