T-SQL Tuesday #137–Notebook Uses


It’s actually my month to host T-SQL Tuesday, and I came up with the notebook idea last year. Aaron Nelson (b | t) sent me a link to a talk he did at a Meetup. It contained some interesting things, showing how you can use notebooks. Worth a watch.

For me, I don’t use notebooks a lot in my work. I had thought about using them to show clients and customers how to use our products, but I hadn’t moved far in that direction. I think too many customers are still using SSMS primarily, and haven’t moved to ADS.

The one place I’ve found them interesting is with Python. When I am sometimes trying to work through an issue, I think it’s easier to run a cell of code and get some results to work through an issue. If I use the REPL, I lose things. In VS Code, which I prefer, I don’t like the split code/terminal. I find that slightly annoying. Notebooks make things easier.

I was doing this recently, as I tried to work on the Advent of Code. I tried different parts of an algorithm in different cells, just to see what the results were. This helped me to work out some logic.

2021-04-13 10_02_06-? Day5.ipynb - Data Analysis - Azure Data Studio

That’s not a great algorithm, and I got better over time, but this was really a good way for me to think about the sections of the problem, putting each on in a separate area.

It’s a little of “teaching myself” within the construct of a notebook. After all, that’s what a lot of data scientists are doing when they start to go through different code items in a notebook.  They can share the notebook with others, which helps to teach.

For me, it’s teaching myself.

Original post (opens in new tab)
View comments in original post (opens in new tab)