I often use Python when creating tools for personal use. The tools I create are generally for automating day-to-day tasks or for fun application projects.
These are usually small projects that I complete in a few days and don’t update much afterward. The considerations are different for larger, publicly released services, but here are two reasons why I frequently choose Python for small tool development.
Reason 1: Python Can Do Almost Anything
When I want to accomplish something, Python often already has a library for it. Here are some examples of Python libraries I've used for personal projects.
Machine Learning
- Python is likely the most well-equipped language for machine learning libraries.
- Although I don’t personally train deep learning models often, I sometimes use scikit-learn or XGBoost to build and apply models.
Image Processing
- I’ve written scripts for managing personal photos.
- Libraries like PIL (Python Imaging Library) and Pillow help me retrieve Exif data or resize images.
Scraping
- I’ve created tools to periodically check information on certain websites.
- You can use simple libraries like Requests, or more comprehensive ones like Scrapy to make scraping even easier.
Cryptocurrency Trading
- I once wanted to use a cryptocurrency exchange API.
- Thanks to the library
ccxt, which allows you to use the APIs of over 100 exchanges with a unified interface, I could achieve what I wanted. - It was very helpful not to have to investigate the API specifications of each exchange and to be able to trade with a consistent interface.
Web Applications
- Sometimes I want to control the above functionalities through a GUI.
- In such cases, I often use Django to run it as a web application.
- I particularly like Django because it provides an admin panel by default, making it easy to manage settings and check data.
Reason 2: It’s Cheap to Run in the Cloud
Since personal tools aren’t used frequently, I want to keep the costs low when running them on a server. Python has long been supported by free cloud platforms, which is another reason I choose it for personal tool development.
Google App Engine (GAE)
- GAE offers a free tier in its standard environment.
- Since it has supported Python since its release in 2008, I’ve often used it for running personal tools.
- It’s also handy that you can set up cron jobs for scheduled execution through the management console.
AWS Lambda
- AWS Lambda was released in 2014, and Python has been supported since October 2015.
- It also offers a free tier, so I sometimes run tools on it nowadays.
- Using the Serverless Framework provides a smooth experience from local development to deployment.
(Depending on the situation, I also run tools on EC2 or Heroku.)
Complaint About Developing with Python
There are some aspects of Python that I find unsatisfactory. In particular, the management of virtual environments and packages tends to be unstable. When I check back after some time, I often find a new method has been introduced or an old method has been deprecated. I’ve used the following tools, but it’s easy to get confused if you don’t understand how to use each one properly. (I’m not sure what the current best practices are.)
- virtualenv
- venv
- pipenv
- pip-tools
- poetry
Conclusion
I’ve listed two reasons why I often use Python for creating personal tools and added one complaint for good measure. I hope this has been helpful.
Top comments (0)