7/7/2023 0 Comments Conda vs miniconda vs anacondaIf you already have Anaconda installed on your computer, you can still install Miniconda on your computer (see additional details in the Setup Miniconda section below). You can always add more Python packages as you need them! To limit the time and space needed for installation (and to minimize potential conflicts between packages), you will use the Miniconda Python distribution to get started with only packages that you need to complete the Python lessons on this website. Installation can take up a lot time and space on your computer Installs Anaconda Navigator, Spyder, and many other tools that may not be needed Installs a long, pre-configured list of Python packages (many of which may not be used) You also learned that the conda package manager allows you to install Python packages on your computer as well as create and manage multiple Python environments, each containing different packages.Īlthough the conda package manager can be installed using either the Miniconda Python distribution or the Anaconda Python distribution, there are key differences between the two distributions: Anaconda Git Bash is used by Windows users to access both Git and Bash in one easy-to-install terminal. In the previous lesson, you learned that Git is a widely used tool for version control that allows you to track and manage changes to your files. Information below is adapted from materials developed by Data Carpentry and the Conda documentation. Open a terminal and test that Bash, Git, and Conda are ready for use on your computer.īefore you start this lesson, be sure that you have a computer with internet access.Install the Miniconda Python distribution.The environment construction method with Miniconda is summarized below.At the end of this activity, you will be able to: I first built the environment with Anaconda, but I couldn't grasp the contents, so I uninstalled it and rebuilt it with Miniconda.Īlthough Anaconda is standard and rich in tools, you end up having to look into the package when you write your own programs.I think it's important that you know what's in it. People who don't like installing unnecessary packages.People who want to know which package they are using.Those who want to start machine learning as soon as possible.People who don't care if there are unnecessary packages.People who do not want to have a hard time building an environment.Which one should build the environment Suitable for Anaconda Installation of python is easy, but necessary packages and execution environment are built individually using conda. The smallest configuration version of Anaconda. Graphical User Interface (GUI): Anaconda Navigator.Integrated Development Environment (IDE): Jupyter, JupyterLab, Spyder, RStudio.Package: numpy, pandas, Matplotlib, Scikit-learn, Tensorflow.If you install Anaconda, you will be able to use packages for scientific calculation and data science together with Python.It also includes "R", a programming language for data science alongside Python, and their comprehensive development environment.Roughly speaking, the following applications are installed. "Python + R language + conda + 1000 or more related packages + execution environment + etc. It's true that Anaconda makes it easy to build an environment, but it also has its disadvantages.Therefore, I compared the characteristics of Anaconda and Miniconda. When it comes to building a machine learning environment with python, many books and sites say that you should use Anaconda for the time being.
0 Comments
Leave a Reply. |