Merge in existing implementations of the correlations. about autocorr HOT 2 OPEN

Hi,

If you are interested, we are also developing a software named dynamix for (one time) correlation function computation. It primarily aims at being simple and fast.
The front-end code is in Python, calling OpenCL/CUDA routines through pyopencl/pycuda.

It features two kinds of correlators : one designed for regular data, one for "sparse" data.
The correlators are unit tested against reference "legacy" implementations, so we are also interested in validating the code on synthetic datasets with analytically known g_2.

The project is still at a early development stage, so we are open to suggestions.

Best regards,

Pierre

from autocorr.

Hi @pierrepaleo. Thanks for reaching out! Your message above must have slipped through my GitHub notifications, and I just now noticed it. That looks very relevant to our interests! I feel we are still in the process of wrapping our minds around what we have and what we need, so it would be great to chat and hear more about what you have and what you need. Would you be interested in having a video chat after the holidays?

from autocorr.