安裝 Miniforge

brew install miniforge

建立 Environment

conda create --name datascience python=3.8 pip conda
conda activate datascience

基本 Python Packages

conda install tensorflow keras pytorch pytest pylint flake8 autopep8 numpy scipy scikit-learn matplotlib seaborn bokeh plotly pandas requests urllib3 bs4 selenium-requests selenium
conda install pytest pylint flake8 autopep8
conda install numpy scipy scikit-learn
conda install matplotlib seaborn bokeh plotly
conda install pandas
conda install tensorflow keras pytorch -c conda-forge
conda install plotnine koalas -c conda-forge
conda install requests selenium selenium-requests urllib3 bs4 (Miniforge 目前還沒有 Selenium)
pip install pyecharts cutecharts

安裝 Jupyter

conda install jupyterlab ipywidgets
conda install jupytext -c conda-forge

安裝 Spyder

brew install --cask spyder 或
conda install spyder spyder-kernels (Miniforge 目前還沒有 Spyder)
conda install spyder-notebook -c spyder-ide

安裝 Python Kernel

% conda install ipykernel
Collecting package metadata (current_repodata.json): done
Solving environment: done

# All requested packages already installed.

% python -m ipykernel install
Installed kernelspec python3 in /usr/local/share/jupyter/kernels/python3

安裝 R Kernel

透過 Terminal 安裝 R Kernel，不要在 RStudio 裡頭安裝：

% R
> install.packages("devtools")
> library("devtools")
> install_github("IRkernel/IRkernel")
> library("IRkernel")
> installspec()
[InstallKernelSpec] Installed kernelspec ir in /Users/kcsu/Library/Jupyter/kernels/ir
> quit()

% jupyter kernelspec list
Available kernels:
  ir         /Users/kcsu/Library/Jupyter/kernels/ir
  python3    /opt/homebrew/Caskroom/miniforge/base/envs/datascience/share/jupyter/kernels/python3

查看 Jupyter Kernel

jupyter kernelspec list

Kernel Switch

透過 Switch Kernel 去 Switch Environment。

Automatic Environment Kernel Detection for Jupyter：

網址： https://github.com/Cadair/jupyter_environment_kernels
目前必須透過 Pip 安裝
使用前必須產生 Notebook Configuration 檔案： jupyter notebook --generate-config

修改 ~/.jupyter/jupyter_notebook_config.py 檔案：

c.NotebookApp.kernel_spec_manager_class = 'environment_kernels.EnvironmentKernelSpecManager'
c.EnvironmentKernelSpecManager.conda_env_dirs=['/usr/local/anaconda3/envs']

執行 Jupyter 的時候，就可以找到所有 Environment 裡頭的所有 Kernel 了：

% jupyter lab --no-browser
...
Starting initial scan of virtual environments...
Found new kernels in environments: conda_learning, conda_anaconda3

Chrome App Mode

執行 jupyter lab --no-browser 指令，然後開啟 Google Chrome 瀏覽器，複製網址與 Token，就可以透過 Chrome App Mode 的方式執行。

如果覺得每次複製貼上很麻煩，可以修改 ~/.jupyter/jupyter_notebook_config.py 檔案 (NotebookApp 或 LabApp 都可以)：

c.NotebookApp.browser = '/Applications/Google\ Chrome.app/Contents/MacOS/Google\ Chrome --incognito --user-data-dir="/tmp/jupyter" --app="%s"'

以後只要執行以下的指令 (記住不要再加上 --no-browser)：

jupyter lab

Google Chrome 瀏覽器就會以 App Mode 的方式，執行 Jupyter Lab。

MonsterSupreme / Blog

Python Learning Environment #4