Jupyter Notebook 🐼

After installing and learning more about the Jupyter Notebook and doing data processing with Python, I think that is the way forward for much of the data analysis I do. Spreadsheets and pivot tables are fine but memorizing a few commands means that I can write code that can be automated and reused unlike things that I do drag and drop in the spreadsheet.

Jupyter is neat because it allows you to edit and execute a few lines of python in your browser, get the results than independently execute additional lines of code. Then display the results as formatted tables, graphs or graphics right in the browser. The more I learn about it, the neater it is.

The more I use Python every day, the better I will get at it and eventually I’ll be able to whip together new analysis without much struggle. So much of Python is optimized towards minimal typing, so it really is the quickest way to process data. And when I’m using Python daily rather than a spreadsheet I’m learning the language and getting better at it.

Some of the ways I think I could use Jupyter:

• Processing the dataset I use from time to time to create graphs of the 7 day rolling average of COVID-19 positivity cases
• Processing data for graphs from the Energy Information Office or any other agency where i can get a CSV or Excel file to manipulate
• Using machine learning looking at weather data over the past 10 or 20 years to make Farmers Almanac style forecasts for fun – here’s your Thanksgiving Weekend forecast in July.
• Process election results programmatically without all those hairy excel formulas, turn into maps and graphs
• Basically any other project that I can imagine that I would normally do with a spreadsheet but probably could do faster and more repeatly with Python