Ecosystem for combining personal data
Nostalgia will help with gathering data from a variety of sources and enable you to combine them easily.
It's much like Home Assistant, providing a platform, but then for connecting data instead of IoT devices.
Afterwards, it will help you filter and visualize the data.
The architecture is as follows. You're looking at the core which contains the code for ingesting sources, installing the backend system and allows you to write scripts using Nostalgia Query Language.
If you want to add your own data that is not supported, please for now contact us directly in either discord or slack and we'll help you get started. You can also look at the open issues to see suggestions for new sources.
Available Data Bindings
Full list of current sources.
- Heartrate (Fitbit, Samsung Watch)
- Sleep (Fitbit, Samsung Watch, SleepCycle)
- Places (Google Timeline)
- Bank Payments (ING)
- Pictures (Google)
- Emails (Gmail)
- App Usage (Google/Android)
- Chat Conversations (WhatsApp, Facebook Messenger)
- Music listening (Google)
- Music identification (Shazam)
- Posts (Reddit, Facebook)
- File Visits (Emacs)
- Annotated Screenshots (Tesseract)
- Web (Nostalgia Chrome Plugin)
- Google Search
- Indoor Positioning (whereami)
- Public Transport (MijnOV)
pip install nostalgia
Follow the instructions for a source of interest to ensure it is loaded
Use the data in an interactive session (run Python) OR run the timeline
Nostalgia Query Language - extending pandas
Given that you have payments, heartrate and google places set up, you could start Python and run:
In : from nostalgia.sources.ing_banking.mijn_ing import Payments payments = Payments.load() payments.by_card\ .last_year\ .in_office_days\ .during_hours(7, 12)\ .by_me()\ .heartrate_above(100)\ .when_at("amsterdam")\ .sum() Out: 7.65 # in euros
This will give the total expenses by card in the last week, but only on work days, at night, when my heart rate is above 80 and I'm in Amsterdam. It combined the generic class functionality, together with data from:
- A Payments provider
- A Location provider
- A Heartrate provider
Please contribute the data sources that others could use as well!