Chris Xiao

> Thanks for the info. I don't have access to an M1 Mac and thus haven't tried to build AMBF on a VM. Did you try the solution to Issue...

> Are you able to make a temp directory and write to it during an HPC job (completely outside an ants.registration call)? Did you mean I create and write to...

> Yes, just trying to rule out potential issues outside of anything antspy-specific. Yes, it can work. I used a numpy array and saved it to the temp directory ```...

> Okay, great. Can you post the command line call and the resulting error message? ``` #!/bin/sh #SBATCH -J label_propagation # job name #SBATCH -N 1 # nodes requested #SBATCH...

> /projects/autoseg-headneck-dl/tmp Script is below: ``` import numpy as np import os import ants def main(): template = ants.image_read('/projects/autoseg-headneck-dl/template/IR_187.nii.gz') target = ants.image_read('/projects/autoseg-headneck-dl/target/IR_191.nii.gz') transform_forward = ants.registration(fixed=template, moving=target, type_of_transform="SyN", syn_metric="demons", reg_iterations=(80, 40,...

The SyN step took a lot of time and it is weird that it is fast on my local laptop

> Are you controlling the number of threads? If not, the job might try to use all the CPUs in the HPC, which can make things much slower. Try setting...

> This sounds to me like resource restrictions on the jobs as being submitted to SLURM (low CPU count and RAM). I suggest contacting your HPC support for help with...

@Riser6 Hi, I have the same question here, did you solve the problem? I use the same code to convert depth map to point cloud but with two different depth...

> > projectionFactor > > How do I find projectionFactor You can use the focal length instead of projectionFactor, here is the modified code: ```py z = depth value you...