Managing Projects And Requirements In Anaconda

Managing projects in anaconda: The Workflow

Often, it is important for certain projects and python files that a certain version of a certain module is installed, or even a certain python version.

You might also have cloned a project via git that has no requirements.txt and you start getting tired of installing each and every module manually.

Installing all modules manually from a file you received can be quite cumbersome

But don’t worry: Here, I will list the steps in order to manage projects in anaconda in a clean, safe way. You will learn how to:

• auto-generate and save requirements.txt for a given file / project
• create a virtual environment in conda (A virtual environment is an isolated workspace where you install your packages separate from the main system installation)
• install all required packages in the requirements.txt in that environment That way, we can avoid unnecessary conflicts between versions of packages and python versions.

---

Using pipreqs to save the requirements.txt of a project or file

1. Install pipreqs

In your terminal, type pip install pipreqs to install pipreqs

2. Save the required modules as requirements.txt of a project

• Using the pipreqs command pipreqs /path/to/project

Or:

• navigate to the folder of the project in the terminal
• type pipreqs . to include all packages

Example usage of pipreqs

Your requirements.txt will be created in the same folder.

---

Create a virutal environment in anaconda and install requirements.txt

Now that we have a requirements.txt file with required modules, we can install them in a virtual environment using conda.

1. Create a virtual environment for your project in conda

• To create a new environment without and install python on it: type conda create --name myenv python

• To create an environment with a specific version of Python: Type conda create -n myenv python=3.9.2 (where 3.9.2 is your desired version)

where myenv is your (custom) environment name

• don’t forget to activate your environment using conda activate myenv

Creating and activating an environment using conda

• You can see all your virtual environments using the command conda info --envs

• You can remove a virtual environment using conda remove --name myenv --all (where –all is optional)

• You can deactivate and get back to your base env using conda deactivate

• See this link for more information about managing environments in conda.

2. Install the requirements in that new environment

• Just use python -m pip install -r requirements.txt to install all the required packages in their versions!

• you can check if you’re in the right pip using the command which pip

• see this link for more information on using pip inside a virtual environment