Installing a PyMC environment

For VS Code

1. Install pixi:
   • Run this in PowerShell:
     • powershell -ExecutionPolicy ByPass -c "irm -useb <https://pixi.sh/install.ps1> | iex"
2. Create the environment in a project directory
   • Download this repository as a zip file and extract it into an empty director
     • This will be your project direcotry
     • Open a command prompt in the new direcotry
     • In the command prompt type:
     • In the cloned repository run pixi install
       • This will install the environment you need for PyMC
3. Set up VS Code
   • Make sure pixi extension to VS Code is installed
   • Type code . to start VS Code
     • Or click Repository -> Open repository in VS Code in GitHub Desktop
     • VS Code should automatically recognize the environment and suggest it as the default Python environment
4. Run the demo files
   • There are two demo files:
     • Straight python: pymc_demo.py
     • Jupyter notebook: pymc_demo.ipynb

For PyCharm

1. Install necessary software
   • Install pixi:
     • Run this in PowerShell:
       • powershell -ExecutionPolicy ByPass -c "irm -useb <https://pixi.sh/install.ps1> | iex"
   • Install mamba
     • Download the installer and run it
2. Create the environment in the project directory
   • Download this repository as a zip file and extract it into an empty directory
     • This will be your project direcotry
   • Open a command prompt in the new direcotry
   • With the current directory being the project direcotry, type into the commmand prompt:
     • pixi project export conda-environment > environment.yml
       • This will create an environment specification file for mamba/conda
         • Mamba / conda are compatible with PyCharm
     • mamba env create -f environment.yml -n project_name
       • Replace project_name with the name for the environment
       • PyCharm expects this to be the name of the directory
     • mamba activate project_name
     • mamba install m2w64-toolchain
       • This installs the compiler that pymc uses
3. Setup PyCharm
   • Open PyCharm and use it to open the project directory
   • Go to Settings (gear bar upper right or Ctrl-Alt-S) -> Project -> Python Interpreter
   • Click add interpreter -> add local interpreter
     • Under Type choose Conda
     • Under Name choose the project_name
       • You may need to type it in
   • Re-read the package list
     • File -> Invalidate caches and then Ivalidate and restart
4. Run the demo files
   • There are two demo files:
     • Straight python: pymc_demo.py
     • Jupyter notebook: pymc_demo.ipynb

For Spyder

There is probably a way to get Spyder to work with a regular install, but I got it to work by installing it into the environment

1. Install pixi:
   • Run this in PowerShell:
     • powershell -ExecutionPolicy ByPass -c "irm -useb <https://pixi.sh/install.ps1> | iex"
2. Create the environment in a project directory
   • Download this repository as a zip file and extract it into an empty director
     • This will be your project direcotry
     • Open a command prompt in the new direcotry
     • In the command prompt type:
     • In the cloned repository run pixi install
       • This will install the environment you need for PyMC
   • Install spyder into the pixi environment
     • pixi add spyder numpy scipy pandas matplotlib sympy cython
3. Run Spyder from within the pixi environment
   • Still in the command prompt and still in the project directory, type:
     • pixi shell
     • spyder
   • This will open a spyder fully setup with the appropriate environment
4. Run the demo files
   • Spyder doesn't support Jupyter notebooks, so you only have the straight Python:
     • pymc_demo.py