Rate This Document
Findability
Accuracy
Completeness
Readability

Installing the Python Dependency Library

The Anaconda virtual environment can prevent the Python environment from being polluted. However, Anaconda has vulnerabilities. Determine whether to use Anaconda based on your requirements. Before installing Anaconda, install the GCC/G++, make, CMake, and Glibc components. After the components are installed, download, install, and initialize Anaconda, create an OmniAdvisor virtual environment, and install required dependencies.

Prerequisites

Before installing the Python dependency library, install GCC/G++, Make, CMake, and glibc on the management node. Table 1 lists the version requirements.
Table 1 Software to be configured before installing the Python dependency library

Item

Version Requirement

Reference

GCC/G++

7.3.0

Link

Make

4.2.1 or later

-

CMake

3.13.4 or later

Link

Glibc

2.25 or later

Link

MariaDB

10.5 or later

-

MySQL

5.7 or later

-

By default MariaDB/MySQL uses port 3306 as the database listen port. Port 3306 is a risky port and may cause security problems. You are advised to change the listen port when deploying MariaDB/MySQL.

Installing the Python Dependency Library

  1. Optional: Install the Anaconda software package on the management node.
    1. Download the Anaconda software package.
      1
      wget -k https://repo.anaconda.com/archive/Anaconda3-2022.05-Linux-aarch64.sh
      
    2. Install Anaconda.
      1
      bash Anaconda3-2022.05-Linux-aarch64.sh
      
    3. Initialize Anaconda.
      1
      source /root/anaconda3/etc/profile.d/conda.sh
      
    4. Establish an Anaconda virtual environment.
      1
      conda create -n OmniAdvisor python=3.10
      
    5. Activate the Anaconda virtual environment.
      1
      conda activate OmniAdvisor
      
  2. Install the dependencies.
    1
    2
    3
    4
    5
    6
    7
    8
    9
    pip install Cython==0.29.24
    pip install numpy==1.22.3
    pip install PyYAML==6.0
    pip install pymysql==1.0.2
    pip install scipy==1.7.3
    pip install scikit-learn==1.0.2
    pip install pandas==1.3.5
    pip install sqlalchemy==2.0.16
    pip install pint==0.22