Hello,
thank you for the software. The cell segmentation look good to us.
We currently are using your software together with sopa, which works well. Still, the duration of a segmentation on one patch, even when having small patches like in your tutorial, is burdensome. Currently, we're facing wall-times per patch of up to 3d and an overall segmentation duration of close to 7d.
Is there a way to increase the speed of the segmentation? I've tried increasing the number of cpus, as I saw that you're using scanpy within your code. Still, that barely accelerated the segmentation. Are there other things worth trying?
Best,
Jonas
Hello,
thank you for the software. The cell segmentation look good to us.
We currently are using your software together with sopa, which works well. Still, the duration of a segmentation on one patch, even when having small patches like in your tutorial, is burdensome. Currently, we're facing wall-times per patch of up to 3d and an overall segmentation duration of close to 7d.
Is there a way to increase the speed of the segmentation? I've tried increasing the number of cpus, as I saw that you're using scanpy within your code. Still, that barely accelerated the segmentation. Are there other things worth trying?
Best,
Jonas