Conda and its benefits beyond python¶
Conda is a powerful, open-source package and environment management system that allows you to manage compilers and packages for various programming languages. Although commonly associated with Python, Conda supports many more languages, making it versatile for diverse computational tasks. Here’s a guide to briefly introduce Conda for a few programming languages. Feel free to jump to the section on the language of your interest! This guide is meant to quickly give you an idea of the available programming languages/compilers, how to import packages, and how to create environments.
Summary of programming languages
Python, Julia, C++ (gcc), FORTRAN (gfortran), R, Java, Perl, Ruby
What is Conda and Why Use It?¶
Conda is useful for:
Managing dependencies: Avoid version conflicts between libraries or packages. One of Conda’s most powerful features is its ability to resolve dependencies automatically when installing multiple packages. When you specify multiple packages in a single conda install command, Conda determines compatible versions of each package to ensure they work together without conflicts. This avoids version mismatch errors, which can occur when packages are installed separately, as they may depend on incompatible versions of shared libraries.
Creating isolated environments: Work on multiple projects with different setups without interference. This also avoids problems when our operating system updates, and with it, the native version of a few programming languages and compilers.
Cross-language compatibility: Manage tools and packages not only for Python, but also for R, Julia, C++, and others.
Dependencies and version conflicts¶
By separately installing packages, i.e. running conda install <package1> then conda install <package2>, you miss the features of Conda where it can resolve the dependencies, e.g. package2 needs a specific version of package1. Instead, it is recommended to use conda install <package1> <package2>, such that you don’t have th manually resolve the version conflicts.
Installing Conda¶
If you don’t have Conda installed, please see our Available software page.
Conda for Python¶
Conda comes with a minimal Python version. You can check the version using python --version and the location with which python, and can install a specific version, e.g. Python 3.5 with conda install python=3.5.
To create a Python environment called python_env:
conda create -n python_env python
You can then write your code, let’s say in a file called example.py, and run it with the command python example.py.
Beyond Python¶
Conda for Julia¶
You can install Julia through Conda, instead of installing directly as presented in the User’s Guide, using the command
conda install -c conda-forge julia
You can then check which version is installed with julia --version and where using which julia.
To create a Julia environment called julia_env:
conda create -n julia_env julia
You can then write your code, let’s say in a file called example.jl, and run it with the command julia example.jl.
Conda for C++¶
Natively, subMIT currently has a C++ compiler, g++. While Conda doesn’t directly install C++ as a standalone compiler, it can install related tools (like GCC [GNU Compiler Collection] or Clang) and libraries for building C++ projects, e.g.
conda install -c conda-forge gcc
To check the version you have installed, use gcc --version and to get its location, use which gcc. To create an environment called cpp_env:
conda create -n cpp_env
You can then write your code, let’s say in a file called example.cpp, and compile it using gcc example.cpp -o example. Finally, you can run it with the command ./example.
Conda for FORTRAN¶
Natively, subMIT currently has a FORTRAN compiler, gfortran. Similarly to C++, Conda can install FORTRAN compilers, such as a specific version of gfortran, through the command:
conda install -c conda-forge gfortran
You can check the version of gfortran through the command gfortran --version, and where it is installed with which gfortran. You can install FORTRAN libraries, e.g.
conda install -c conda-forge lapack blas fftw
Create an environment:
conda create -n fortran_env gfortran
You can then write your code, let’s say in a file called example.f90. Compile your code with gfortran example.f90 -o example, and run with ./example.
Conda for R¶
Conda can also install R
conda install -c r r-base
Use R --version to determine the version of the language, and which R for its location.
To create an environment for R, use
conda create -n r_env r-base
To run a script called example.R in R, use Rscript example.R.
Conda for Java¶
Java is also natively installed on subMIT. If you wish a different version, you can for example install it using
conda install -c conda-forge openjdk
Use java --version to determine the version and which java for its location.
Some, but not all, Java-related libraries are available via Conda, e.g.
conda install -c conda-forge java-jline
Conda for Perl¶
Perl is also natively installed on subMIT. If you wish a different version, you can for example install it using
conda install -c conda-forge perl
perl --version will give the version you have installed, and which perl, its location.
To import Perl libraries, such as perl-dbi, run
conda install -c conda-forge perl-dbi
Conda for Ruby¶
Ruby is not natively installed on subMIT. You can install it through
conda install -c conda-forge ruby
ruby --version will give you the version you have installed, and which ruby its location.
Conda’s ability to import Ruby packages is limited. You can manage Ruby gems indirectly or use Ruby libraries available through Conda, e.g.
conda install -c conda-forge ruby-rails
Ruby environments can also be created with Conda.
How about pip?¶
pip and Conda are package management tools commonly used for Python. The main features of pip are:
Python-focused pip is a package manager specifically for Python
Dependencies pip does not perform dependency resolution like Conda. pip will install the latest version of each package, without checking if some packages require an earlier version to be compatible.
Environments pip’s virtual environments,
venv, can be created and activated usingpython -m venv my_env
Exporting environment with both Conda and pip, we can export an environment to share it with other users. The commands are
conda env export > environment.yml pip freeze > requirements.txt
These environments can then be recreated by other users by running
conda env create -f environments.yml pip install -r requirements.txt
