docs/lol-paper/scratch/flashlol.md

In neurodata/fselect: Linear Optimal Low-Rank Projection

STEPS

1. AMI Type

ami-0698bcaf8bd9ef56d

Connect:

ssh -i /path/to/pemkey ubuntu@<ip>

1. Download FlashLOL docker:

Dataset="SWU4"
docker pull ericw95/flashlol:0.0.3

1. Format and Drives:

# format drives
sudo mkfs.ext3 /dev/xvdb

# mount drives
sudo mkdir -p /mnt/ssd1

sudo mount /dev/xvdb /mnt/ssd1

sudo mkdir -p /mnt/ssd1/dwi

1. Download Data data and make available:

screen -R
sudo aws s3 cp --no-sign-request --recursive s3://mrneurodata/data/$Dataset/ndmg_0-0-48/reg_dti/ /mnt/ssd1/dwi
sudo aws s3 cp --no-sign-request --recursive s3://mrneurodata/data/$Dataset/ndmg_0-0-48/reg_dti/ /mnt/ssd1/
sudo chmod -R 777 /mnt/ssd1

1. Clone repo:

mkdir ~/Documents
cd ~/Documents
git clone https://github.com/neurodata/lol.git

1. Launch Docker:

docker run -ti --entrypoint /bin/bash -v /<path>/<to>/<repo>/lol:/lol -v /mnt/ssd1:/brains ericw95/flashlol:0.0.1

1. Run scripts:

cd /lol/docs/lol-paper/scratch/

neurodata/fselect documentation built on March 6, 2021, 12:54 p.m.