https://github.com/tindzk/nix-ds
Nix for Data Science
https://github.com/tindzk/nix-ds
data-science nix python
Last synced: 11 months ago
JSON representation
Nix for Data Science
- Host: GitHub
- URL: https://github.com/tindzk/nix-ds
- Owner: tindzk
- Created: 2016-11-16T14:51:45.000Z (over 9 years ago)
- Default Branch: master
- Last Pushed: 2016-11-29T19:08:42.000Z (over 9 years ago)
- Last Synced: 2025-03-15T23:45:48.409Z (over 1 year ago)
- Topics: data-science, nix, python
- Language: Nix
- Size: 3.91 KB
- Stars: 7
- Watchers: 2
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Nix for Data Science
As Ubuntu's LTS releases usually do not ship the latest versions, it becomes challenging to set up a clean environment for Data Science. Nix provides us with an easy-to-use alternative without needing to interfere with the host system.
Here, we build a Python 3.5 environment with packages commonly used for Data Science. The packages are defined in `default.nix` instead of using Python's package manager (`pip`).
## Nix
First off, install Nix:
```bash
$ bash <(curl https://nixos.org/nix/install)
$ source ~/.nix-profile/etc/profile.d/nix.sh
```
## Building the environment
Clone this repository, change the working directory to it and run `nix-build`:
```bash
$ git clone git@github.com:tindzk/nix-ds.git
$ cd nix-ds
$ nix-build
```
Now, instead of using the host's Python tools, you can use the ones from `result/bin/`:
```bash
$ result/bin/python3
$ result/bin/ipython notebook
```
## Upgrading environment
The environment can be easily upgraded:
```bash
$ nix-channel --update
$ nix-env -u
```
## See also
* [Getting started with Nix package manager](https://www.domenkozar.com/2014/01/02/getting-started-with-nix-package-manager/)
* [Nix for Python developers](http://datakurre.pandala.org/2015/10/nix-for-python-developers.html)