I wanted to reach out with a couple of questions about best practices for working with TNG cluster data via SCIDA. I’m working in a JupyterLab environment with access to TNG and can load simulations, snapshots, and clusters; however, the read times are proving challenging.
For example, when I load a single cluster at z=0, the SCIDA group load often takes close to 10 minutes. I’ve seen some discussion on GitHub about caching times (which I assume is related), but I wanted to check in directly to learn how you typically handle this in your own workflows. My goal is to be able to iterate over clusters in the simulation in a reasonably efficient manner, but at present this step is a significant bottleneck.
Thank you for any advice or pointers you can share!
Dylan Nelson
8 Sep
This 10 minute time should be once only right, and then it should be essentially zero thereafter?
I wanted to reach out with a couple of questions about best practices for working with TNG cluster data via SCIDA. I’m working in a JupyterLab environment with access to TNG and can load simulations, snapshots, and clusters; however, the read times are proving challenging.
For example, when I load a single cluster at z=0, the SCIDA group load often takes close to 10 minutes. I’ve seen some discussion on GitHub about caching times (which I assume is related), but I wanted to check in directly to learn how you typically handle this in your own workflows. My goal is to be able to iterate over clusters in the simulation in a reasonably efficient manner, but at present this step is a significant bottleneck.
Thank you for any advice or pointers you can share!
This 10 minute time should be once only right, and then it should be essentially zero thereafter?
The discussion is: https://github.com/cbyrohl/scida/issues/195
(Probably best to discuss scida related issues on its github).