Data sources

On the sources for my report data:

Time tracking data

Recorded manually - details on the methods page.

Finance/spending data

I track spending and do budgeting in GnuCash. GnuCash has a reasonable data format (XML) and reasonable export options (CSV, SQL), but I used the built-in reporting tools to summarize my food spending instead of reinventing the wheel.

Most of the raw data comes from my banks (Simple, Discover). I also have an Apple Card, which doesn’t support any export formats whatsoever. I retyped every Apple Card transaction from 2019 from their PDF statements into GnuCash. After I did that, I found this Apple Card statement converter which does the conversion automatically. Cool cool cool

To-do item performance

I track my to-do items in Things.

I have been using Things for almost 10 years now, and I like a lot of things about it, but Cultured Code has a haughty and dismissive attitude towards the concept of data integration. There are no supported methods for programmatically reading or writing task data to/from Things (except for a custom URL scheme, which I don’t count as a real method).

I’ve been working on building a library to read the Things cloud sync data for some time – which would give me the ability to recreate the state of my to-do list at any point in time – but it’s proven to be a challenging project.

In the meantime, I did these analyses by reading the Things.app SQLite database on the Mac app, which is pretty straightforward to interpret.

Messaging data (iMessage, SMS)

The macOS Messages.app stores message data in another SQLite database, chat.db. There are several solid resources for analyzing the contents of this database out there on the interbutt.

I’ve had a Mac with Messages.app since iMessage came out, and I don’t delete messages, so I have a pretty comprehensive database (653MB, not including images!). It’s a pretty neat dataset.

Spotify activity

Spotify has a GPDR data download page. They say it can take 30 days (!) to get the data, but it was closer to a week for me.

Miles driven

I’ve been logging all of the fuel-ups for all of my vehicles using Fuelly for the last… nine years. Yikes.

Fuelly has reasonable CSV export options so this was pretty easy.

Health data

I consolidate all of my health data in the iOS Health app, which has a built-in XML export function.

Most of my weight data comes from a eufy smart scale, which appears to have no model number. It works fine.

Blood pressure data comes from a Greater Goods Bluetooth Blood Pressure Monitor, again apparently too advanced to have a model number. It is also “fine”.