- Install Python interpreter
- Install PyCharm community IDE
- Install python packages
Install Python interpreter
Python interpreter is the program that will run our actual Python code. The default downloaded interpreter is also called “base interpreter”, and for each project it’s recommended to open a new virtual environment (“venv”) with its own “project interpreter”.
Download Python3.6.8 64bit installer .exe from: https://www.python.org/downloads/release/python-368/ (bottom of page)
Open installer -> check the “add Python 3.X to Path” -> click “Install Now” -> press “next” if needed until finish.
(it is preferable to download the python base interpreter to the installer’s suggested place if you don’t have default admin privileges on the computer [i.e.: C:\Users\chech\AppData\Local\Programs\Python])
Install PyCharm community IDE
- Download PyCharm community from: https://www.jetbrains.com/pycharm/download/#section=windows.
- Press “next” until finished.
Open PyCharm -> “create a new project”
Choose the project location -> open “project interpreter …” sub panel -> “new environment…” -> choose location for new environment (I like to put it next to the base interpreter) -> choose base interpreter (the one that was downloaded in section 1) -> click “create”.
First test script
Open a new python file called “test.py”
Choose Run -> run…
You should see the output “hello world” on the cmd.
Install python packages
Packages are the main way to get nontrivial functions that we are going to use, like:
- Numpy: for matrix and vector operations
- Matplotlib: for graph visualization
- Cv2: for image processing tasks
- Tensorflow: for neural networks
We can download all this with the “pip” program (which came with the python interpreter)- the python package manager.
Open the terminal view in pycharm (in the project you are working on)
Check that you are indeed in the right “venv” (will appear in brackets in the cmd), and install all packages with
pip install <package_name>:
pip install matplotlib(will also install numpy…)
pip install tensorflow(a GPU enabled package also exists, try installing it yourself if you wish- it’s not for beginners)
pip install opencv-python
Test package installation
Change test.py to this code:
import tensorflow as tf import numpy as np import matplotlib.pyplot as plt import cv2 print("hello world")
If all installation went well, you should see “hello world” in the cmd.