Environment_Setup_Guide - Lin-Brain-Lab/machine_learning GitHub Wiki

Environment Setup Guide

1. IDE Setup (VS Code)

This section is for users who do not have access to the Sunnybrook VPN or on-site network but want to use an IDE (Visual Studio Code) to connect to the lin1 server and access compute resources.

1. Install the Correct Version of VS Code

Starting with VS Code 1.86, the Remote-SSH extension requires glibc 2.28 or later. However, the lin1 server only has glibc 2.17.

To avoid compatibility issues:

Download and install VS Code 1.85: https://code.visualstudio.com/updates/v1_85
Downgrade the Remote-SSH extension to a version compatible with VS Code 1.85.

2. Configure SSH

Open the Command Palette:

macOS: Cmd + Shift + P
Windows/Linux: Ctrl + Shift + P

Type: Remote-SSH: Open SSH Configuration File Select the config file to edit (usually ~/.ssh/config).

Add the following configuration:

Host robin
  HostName 142.76.1.189
  User username

Host lin1
  HostName 142.76.29.235
  User username
  ProxyJump robin

3. Connect to lin1

Open the Command Palette again:

macOS: Cmd + Shift + P
Windows/Linux: Ctrl + Shift + P

Type: Remote-SSH: Connect to Host Select lin1 from the list.

You will be prompted to enter your passphrase twice:

Once for 142.76.1.189 (robin)
Once for 142.76.29.235 (lin1)

If successful, the lower-left corner of VS Code should display: SSH: lin1.

2. Conda Environment Setup

1. Go to Your Home Directory

cd ~
pwd   # should print something like /home/username

2. Download the Anaconda Installer

For example, for 64-bit Linux:

wget https://repo.anaconda.com/archive/Anaconda3-2025.06-0-Linux-x86_64.sh

(If you prefer a lighter version without bundled packages, you can install Miniconda instead.)

3. Run the Installer

bash Anaconda3-2025.06-0-Linux-x86_64.sh

During installation:

When asked for the install location, press Enter to accept the default:
```
/home/username/anaconda3
```
When asked whether to initialize Anaconda by modifying your shell profile (e.g., ~/.bashrc), type yes.

4. Reload Your Shell Configuration

source ~/.bashrc

This ensures the conda command is available immediately in your current shell.

5. Verify the Installation

conda --version
which conda

Expected output:

/home/username/anaconda3/bin/conda

6. Create and Activate an Environment

conda create -n myenv python=3.8
conda activate myenv