conda_setup.md

Anaconda Setup Guide for FluxPipeline

This guide provides step-by-step instructions for setting up the FluxPipeline project using Anaconda. Anaconda helps manage Python environments and dependencies efficiently.

Table of Contents

• Anaconda Setup Guide for FluxPipeline
  • Table of Contents
  • Prerequisites
  • Installing Anaconda
    • Windows Installation
    • macOS Installation
    • Linux Installation
  • Creating the Project Environment
    • Step 1: Open Anaconda Prompt or Terminal
    • Step 2: Create a New Environment
    • Step 3: Activate the Environment
  • Installing Dependencies
    • Option 1: CPU-only Installation
    • Option 2: NVIDIA GPUs (CUDA support)
    • Option 3: AMD GPUs (ROCm support)
    • Option 4: Intel GPUs (OneAPI support)
  • Troubleshooting Common Issues
    • Issue 1: Conda Command Not Found
    • Issue 2: Conflicts During Dependency Installation
    • Issue 3: Environment Activation Fails
  • Examples of Environment Commands
    • Listing All Environments
    • Deactivating an Environment
    • Removing an Environment
  • Additional Tips
    • Updating Conda
    • Creating an Environment from a YAML File
  • Further Assistance

---

Prerequisites

• Internet Connection: Required to download Anaconda and project dependencies.
• Administrator Rights: May be needed for installation on some systems.

Installing Anaconda

Download and install Anaconda appropriate for your operating system.

Windows Installation

1. Download Installer:

   • Go to the Anaconda Distribution page.
   • Click on Download for Windows.

2. Run the Installer:

   • Double-click the downloaded .exe file.
   • Follow the prompts in the setup wizard.

3. Add Anaconda to PATH (Optional but Recommended):

   • During installation, check the option Add Anaconda to my PATH environment variable.

   Explanation: This allows you to use conda commands from any command prompt.

4. Complete Installation:

   • Click Install and wait for the process to complete.

macOS Installation

1. Download Installer:

   • Go to the Anaconda Distribution page.
   • Click on Download for macOS.

2. Run the Installer:

   • Open the downloaded .pkg file.
   • Follow the installation prompts.

3. Add Anaconda to PATH:

   • During installation, you may need to run conda init in your terminal.

4. Complete Installation:

   • Finish the setup and close the installer.

Linux Installation

1. Download Installer:

   • Go to the Anaconda Distribution page.
   • Download the Linux installer (e.g., Anaconda3-2023.07-Linux-x86_64.sh).

2. Run the Installer:

   bash Anaconda3-2023.07-Linux-x86_64.sh

3. Follow the Prompts:

   • Review the license agreement.
   • Choose the installation location (default is usually fine).

4. Initialize Conda:

   • When prompted, type yes to initialize Conda.

Creating the Project Environment

Step 1: Open Anaconda Prompt or Terminal

• Windows: Open Anaconda Prompt from the Start Menu.
• macOS/Linux: Open your terminal application.

Step 2: Create a New Environment

Create a new environment named flux with Python 3.11:

conda create -n flux python=3.11 -y

Explanation: This command creates a new environment called flux with the specified Python version.

Step 3: Activate the Environment

Activate the newly created environment:

conda activate flux

Example Output:

(flux) C:\Users\YourName>

Installing Dependencies

With the environment activated, install the required dependencies.

Option 1: CPU-only Installation

pip install -r requirements.txt

Explanation: Installs all packages listed in requirements.txt for CPU-only operation.

Option 2: NVIDIA GPUs (CUDA support)

Ensure you have the NVIDIA CUDA Toolkit installed.

pip install -r requirements_cuda.txt

Explanation: Installs GPU-accelerated packages suitable for NVIDIA GPUs.

Option 3: AMD GPUs (ROCm support)

Ensure you have the AMD ROCm platform installed.

pip install -r requirements_rocm.txt

Explanation: Installs GPU-accelerated packages suitable for AMD GPUs.

Option 4: Intel GPUs (OneAPI support)

Ensure you have the Intel oneAPI Toolkit installed.

pip install -r requirements_intel.txt

Explanation: Installs GPU-accelerated packages suitable for Intel GPUs.

Troubleshooting Common Issues

Issue 1: Conda Command Not Found

Solution:

• Ensure Anaconda is added to your PATH.
• Alternatively, use the Anaconda Prompt (Windows) or run source ~/anaconda3/bin/activate (Linux/macOS).

Issue 2: Conflicts During Dependency Installation

Solution:

• Update Conda and Pip:

  conda update -n base -c defaults conda
  pip install --upgrade pip

• Try creating a new environment.

• Check for version conflicts in the requirements files.

Issue 3: Environment Activation Fails

Solution:

• Ensure the environment was created successfully.

• Check the list of environments:

  conda env list

• If the environment doesn't exist, recreate it.

Examples of Environment Commands

Listing All Environments

conda env list

Example Output:

# conda environments:
#
base                  *  /home/yourname/anaconda3
flux                     /home/yourname/anaconda3/envs/flux

Deactivating an Environment

conda deactivate

Explanation: Returns you to the base environment.

Removing an Environment

conda remove -n flux --all

Warning: This deletes the flux environment completely.

Additional Tips

Updating Conda

Keep Conda up to date to avoid issues:

conda update -n base -c defaults conda

Creating an Environment from a YAML File

If a environment.yml file is provided:

conda env create -f environment.yml

Explanation: Creates an environment with all specified dependencies.

Further Assistance

• Conda Documentation: Managing Conda Environments
• Anaconda Support: Anaconda Knowledge Base
• Contact: For project-specific issues, refer to the Contact section in the README.