Open florgf88 opened 1 month ago
We are about to release the new version on the main
branch. New installation instructions are now on the README page. We recommend using a Python virtualenv
as we found this deals with having cupy
installed alongside Torch better. Feel free to post the specific issue you are having!
Hi, many thanks for the response. I managed to install version 1.7.3, which is working very well. I want to apply this pipeline to old recordings that are pretty noisy. Then, all detected channels are either dead or noisy, raising an error when computing the covariance matrix. I was thinking of computing some preprocessing steps outside the pipeline. Is there any way to skip specific preprocessing steps to avoid duplication (e.g. temporal channel alignment)? Many thanks!
Hello! I found some time ago that there was a new release of the Ibl sorter with a better drift motion correction and improvements in the spike holes issue. I was wondering if this version is available to use. I tried to set the ibl_prod_dev version in a fresh conda environment and I found some issues related to the dartSort and torch. Many thanks!