Installing Python on your Windows 64 bit machine

Run this to install python on your PC, 64 bit. To check if you already have python installed, see this link.

If you end up multiple "pythons" installed on your computer, telling the computer where to look for python packages may be tricky - so it's best to just have one single python, and I recommend Miniconda3.

If you already have Anaconda3 already installed and you wish to continue with this tutorial, navigate to ProgramData/Anaconda3 and run the uninstall .exe file.

Step 1

Please make a folder called code under this path, using your username: /Users/my_user_name/Documents/code. This is where we'll keep our repositories.

Download these:

• miniconda3 for python3 64 bits. Barebones python, similar to Anaconda, but without the extra packages.
• GIT for windows. This is a file version management system. Comes with gitbash command prompt. You can also use the Anaconda prompt powershell.
• pip is used to install most python packages, for instance pip install numpy. Right-click this link and select save link as. Save this get-pip.py file under the code directory you just created above.

Step 2

Install python3 64 bits by running the miniconda3 .exe file. ⚠️ When prompted, make sure to install for all users (admin)". ⚠️ This will ensure miniconda3 is installed at C:\ProgramData\miniconda3. Then, open up the Windows environment variables at "Edit the system environment variables" in Control Panel. Click on Environment Variables then select the Path variable under the System Variables list and click Edit. Add a new environment variables with this path: C:\ProgramData\miniconda3 and C:\ProgramData\miniconda3\Scripts

Step 3

Many python packages are installed using makeand Makefiles. To do this, you can install chocolatey. Here's how:

First, login to GitHub (make sure you have made a Github account - with UCSB email, you can make a Pro version)

Next, open windows Command Prompt as administrator. You can do this by typing cmd in the windows search bar and right click to Run as administrator.

Finally, copy-paste this entire line into the terminal:

@"%SystemRoot%\System32\WindowsPowerShell\v1.0\powershell.exe" -NoProfile -InputFormat None -ExecutionPolicy Bypass -Command "[System.Net.ServicePointManager]::SecurityProtocol = 3072; iex ((New-Object System.Net.WebClient).DownloadString('https://chocolatey.org/install.ps1'))" && SET "PATH=%PATH%;%ALLUSERSPROFILE%\chocolatey\bin"

then run

choco install make

If there are issues with this step, you may need to reboot the computer and try it again.

Step 4

Set up pip. This is a platform to easily install packages, example pip install numpy install numpy, a powerful math processing package. Navigate to the location of the get-pip.py file you downloaded above (e.g. cd /yourpath). Then run,

python get-pip.py

Pip is now installed! Now you can run pip install numpy and other useful packages. These include:

• pip install jupyterlab
• pip install pandas
• pip install scipy
• pip install numpy
• pip install matplotlib

If you'd like to see the power of jupyterlab, check out this link. More information about Jupyter. Jupyter offers notebooks that can run python. You evaluate pieces of code sequentially, similar to Mathematica. To open JupyterLab, open GitBash terminal, navigate to your directory of choice, then run jupyter-lab in the terminal. A JupyterLab window should open in your Google Chrome.

Note

⚠️ One known issue is a permissions error when using pip on Windows. To fix this, navigate to the root folder for Miniconda3, which in many cases is under C:\ProgramData\Miniconda3. Right click then Properties, then select the Security tab. For both Administrator and Users, select Full Control of the folder.

Cloning into a repository from Github

You are ready to use any github repo by following the install steps for that repo page. Usually these steps involve the following:

1. Go to the repo github page, for example lightlab
2. In the upper right hand corner, click the green button and copy the URL

3. In your favorite terminal, type:

git clone <the_url_you_copied>
cd <name_of_the_repo>

4. Lastly, some repositories have a last step to finish the installation, basically just follow those instructions. For example, sometimes you have to run make install which if you have chocalatey set up then you can do that. Others have a setup.py file to run to finish the install. Otherwise another useful thing to run is pip install -r requirements.txt to make sure you have all of the repo required packages installed, but again a good repository will walk you through the installation instructions.