I recently came across this site: Memair.com. It seems to have been around as a concept at least for a couple of years but I’d never seen it before this current release. Haven’t yet tried it, but I think it’s intriguing. I know that Open Humans has been using Jupyter Notebooks for their “Personal Data Notebooks” project. Many of us here on this forum have had offline discussions about the prospect of being able to share frameworks for data analysis, and it’s very interesting to see that it’s had some attention from its developer Gregory Clarke recently.
I hadn’t seen that, but that’s a neat use if MyBinder (fun fact, Tim, who helped us setting up the JupyterHub integration on Open Humans is one of the MyBinder contributors)!
It looks like they are doing pretty much the same approach that Open Humans does for getting access: Generating an access_token that can be used to request your personal data from memair, so you can just run a notebook by modifying that token. The main difference is that by running our own JupyterHub on Open Humans we can “magically” set the right token behind the scenes, so you don’t have to fiddle with that yourself.
Awesome to see that someone else is working on this too!
Thanks for starting the conversation @Agaricus. I’ve slowly been hacking away at this project and would love some input from current self quantifiers. My intent is to make a powerful api to make collecting data automated and easy. I’m interested specifically in what data types people are interested in. Currently Memair supports
Biometrics
Digital Activities
Emotions
Journals
Locations
Physical Activities
Next I’ll be adding support for ingestion and I’m open to ideas.
Hi Greg. Definitely glad to have you here on the forum. I think Bastian can testify - and I know from my own conversations through many years with people creating QS tools - that the goal of making data collection “automated and easy” via APIs is very challenging. But if you post updates about your progress here you will probably find a knowledgeable and experienced group.
Yeah, keeping those API integrations up to date is a big task (says the person who wasted a week and updating a tool that uses the API of Moves just a few months ago )
The main difference is that by running our own JupyterHub on Open Humans we can “magically” set the right token behind the scenes, so you don’t have to fiddle with that yourself.
)