Setting up Jupyter Lab in a Python 3.9 venv environment
Introduction
――I always forget about the setup, so I will expose the memo.
environment
- MacOSX --Python installation uses pyenv
- Python 3.9.0 --nodejs is included (necessary for extension installation)
Installation
Creating a venv environment
$ mkdir jupyterlab-sample
$ ce jupyterlab-sample
$ python3 --version
Python 3.9.0
$ python3 -m venv .venv
Main unit pip install
$ .venv/bin/pip3 install jupyterlab
$ .venv/bin/jupyter-lab --version
2.2.9
Start-up
.venv/bin/jupyter-lab
v Start like this
Add package
OK if you install using pip of venv
$ .venv/bin/pip3 install numpy
$ .venv/bin/pip3 install matplotlib
Make it a dark theme
Why ...
vim style key binding
matplotlib settings
Frequently set settings for matplotlibrc.
You can check the path of matplotlibrc currently used by the following method.
from matplotlib import matplotlib_fname
matplotlib_fname()
If it is displayed as ^, it is possible to overwrite the setting by placing a separate matplotlibrc in the working directory with the built-in setting.
Display in Japanese
If you use it normally, Japanese will become tofu, so specify the ttc font file for Mac.
I added Hiragino Maru Gothic Pro to the prefix to the original font.sans-serif as follows.
matplotlibrc
font.sans-serif: Hiragino Maru Gothic Pro, DejaVu Sans, Bitstream Vera Sans, Computer Modern Sans Serif, Lucida Grande, Verdana, Geneva, Lucid, Arial, Helvetica, Avant Garde, sans-serif
I want to make matplotlib a cool graph
Let's use seaborn. https://github.com/mwaskom/seaborn
Installation
.venv/bin/pip3 install seaborn
Can be used by importing
import seaborn as sns
Set the theme and font to your liking
seaborn.set_style("dark")
sns.set(font='Hiragino Maru Gothic Pro')
- (As of December 2020) The trouble is that even though I specified the Japanese font with
matplotlibrc, seaborn seems to be hard-coded, and Japanese is not displayed properly unless I specify the font individually. ....
from now on
There is an LSP for jupyterlab, so I want to try it properly
https://github.com/krassowski/jupyterlab-lsp
Recommended Posts