How to reproduce this experiment¶
Before reproducing this experiment, make sure that:
You have installed the
brainlitpackage [Documentation]. Currently, you need to install the package from source to execute these codes.
You have installed PyTorch [Documentation].
This experiment does not require a GPU (CUDA is not needed for PyTorch).
N.B. if you are using conda, make sure that both
brainlit and PyTorch are installed within the same environment.
Now, follow these steps to reproduce the results of the experiment:
Download segments, which are stored in the publicly available S3 bucket open-neurodata
This script will prepare the experiment folder scaffolding
axon_geometry ├── data │ ├── brain1 │ │ ├── segments_swc │ │ └── trace_data │ └── brain2 │ ├── segments_swc │ └── trace_data ├── figures │ ... etc.
and download data from S3 (no credentials are required).
Compute and save trace analysis data
Run any of the notebooks in the
notebooksfolder, which will save the results in the