Windows Local Python Environment Installation Guide

Within this guide, we will walk you through the steps to setup an environment on your local computer that is running Windows.

Installing Miniconda 3

First, we need to download Miniconda 3 from: https://docs.conda.io/en/latest/miniconda.html

1. Choose the “Miniconda3 Windows 64-bit” option.
2. Open the downloaded installer.
3. Follow the steps to completion.
   • Note: Make sure to select the checkbox to “Add Miniconda to PATH”
4. Open Command Prompt
   • By typing in “search” in the lower left-hand corner cmd.
   • Select Run as Administrator to open the terminal.
5. Ensure that conda was installed correctly and is up-to-date by typing:

conda update -n base -c defaults conda

Installing the inmas Bootcamp environment

With the software prerequisites done, we move onto to setting up a custom conda environment for the bootcamp.

We will use a requirements.txt to pre-specify python packages required for the workshop. You can download a copy by clicking:

Requirements: inmas.yaml

The contents of the file is viewable at the end of the post.

If on Windows, please:

1. Go to the following URL: https://inmas-training.github.io/inmas-bootcamp/inmas.yaml
2. Click “File -> Save Webpage as” and Press “Save”
3. Open Command Prompt
   • Version 1:
     • By typing in “search” in the lower left-hand corner cmd.
     • Select Run as Administrator to open the terminal.
   • Version 2:
     • Or, use Windows Key + R to open Run box.
     • Inside of the box, type: cmd
     • Press Ctrl + Shift + Enter to open an administrator command line.

In Command Prompt, we will navigate to where the configuration file is saved by typing:

cd %HOMEPATH%\Downloads

Next, we’ll use the configuration file downloaded onto the desktop to create the course environment.

# Create an environment called "inmas" with packages installed.
conda env create -f inmas.yaml

Python Environment Details

Contents of the Inmas environment file: inmas.yaml

name: inmas
channels:
- defaults
- pytorch
- conda-forge
dependencies:
- python=3.9
- pip>=21.0
- tensorflow=2.4
- pytorch=1.9
- torchvision
- torchaudio
- scikit-learn
- scipy
- statsmodels
- typing
- requests
- pyyaml
- pandas
- pandas-datareader
- imageio
- pillow
- librosa
- h5py
- tqdm
- matplotlib
- seaborn
- bokeh
- jupyter
- boto3
- pip:
  - gym
  - gradio