Data management platforms

[davclark]

An email to Evan Savage:

I was e-introduced to Dawn some time ago by my colleague (Joe Dumit at UC Davis). Since then, I've become very interested in the data sense project, and I'd very much appreciate if someone could come and participate in an interactive seminar on campus in the fall.

I'm offering this course via the I School, currently scheduled for Thursdays from noon-2pm. The goal would be to have ~1 hour of focused discussion, after which busy folks can leave, and the people who are really engaged have the space to dig in:

http://hackingmeasurement.berkeley.edu

​We'll also be ​running the BIDS Collaborative in parallel with the seminar series, in which folks will be hacking on data science projects. This fall, we particularly invite data collection and management projects, and using data sense (or comparing it with other platforms) would be a great fit.

We'll have a focused practicum Tuesdays 3:30-?? in BIDS, and you would be most welcome to join in helping us apply data sense to one or more projects!

[dnafus]

Hi Dav, I/we'd love to come! I'm trying to get a few trips down to the bay area booked-- could be week of August 24, and potentially w/o Sept 7 (or maybe the week after). If all else fails, I can make a day trip. When does term start for you? When would you need a date locked down?

Dawn

[davclark]

Thanks for the quick reply Dawn! I have done some thinking, and I'm going to delist the course, and revert to a "Social Science Matrix" seminar series, which is what we did in the Spring. It was very productive (and I'm realizing I don't have the bandwidth to organize a proper course every week for two hours).

That said, I'll still be organizing seminars, though I'll now be a bit more flexible on time. The week of Aug 24 is too soon, though. I'm traveling and will be teaching Python the week I get back.

So let's think about doing something the week of Sept 7? If the Thursday at noon slot works, we can do that, but like I said - we can be flexible. I want to make sure we'll get a good set of people in the room. We'll be doing a recurring collaborative project hack sessions on Tuesdays, for example - that would be another good time to visit us.

[dnafus]

I hear you (i.e., not trying to scramble to put a new course together in a few weeks time. Could we say tentatively Tuesday Sept 22, for a project 'hack' session? I might want to move this up to the week of the 7th if I'm indeed going that week, but it's not clear yet whereas I definitely will be around on Sept 24th (an event that conflicts with your class), so could potentially extend the trip.

[candu]

I'll be out-of-town Sept 7, but Sept 22 works fine for me. (I'm in the Berkeley/Oakland area, so I can drop by most anytime given a week's notice or so. That said, students would probably get more out of having us both there, given our complementary perspectives.)

[candu]

Did we settle on a date for this? The more advance notice I have, the easier it is to block off the time on my calendar.

[davclark]

Sorry - I've been meaning to re-factor the syllabus page, but let's do Sept 22 for you all if that still works.

Thanks for checking in!

D

On Wed, Aug 19, 2015 at 2:56 PM, Evan Savage notifications@github.com wrote:

Did we settle on a date for this? The more advance notice I have, the easier it is to block off the time on my calendar.

— Reply to this email directly or view it on GitHub https://github.com/hackingmeasurement/hackingmeasurement.github.io/issues/22#issuecomment-132800265 .

Dav Clark Data Scientist Berkeley D-Lab + BIDS bead.glass 917-544-8408

[dnafus]

That works for me. I’ll block it off too.

From: Dav Clark [mailto:notifications@github.com] Sent: Wednesday, August 19, 2015 3:02 PM To: hackingmeasurement/hackingmeasurement.github.io Cc: Nafus, Dawn Subject: Re: [hackingmeasurement.github.io] Data management platforms (#22)

Sorry - I've been meaning to re-factor the syllabus page, but let's do Sept 22 for you all if that still works.

Thanks for checking in!

D

On Wed, Aug 19, 2015 at 2:56 PM, Evan Savage notifications@github.com<mailto:notifications@github.com> wrote:

Did we settle on a date for this? The more advance notice I have, the easier it is to block off the time on my calendar.

— Reply to this email directly or view it on GitHub https://github.com/hackingmeasurement/hackingmeasurement.github.io/issues/22#issuecomment-132800265 .

Dav Clark Data Scientist Berkeley D-Lab + BIDS bead.glass 917-544-8408

— Reply to this email directly or view it on GitHubhttps://github.com/hackingmeasurement/hackingmeasurement.github.io/issues/22#issuecomment-132801386.

[davclark]

Ok - I've put up a stub that now shows up here:

http://hackingmeasurement.berkeley.edu/syllabus/

It is easy to edit by clicking the little pencil icon (assuming you're logged into GitHub) here:

https://github.com/hackingmeasurement/hackingmeasurement.github.io/blob/master/syllabus/_posts/2015-09-22-Intel-Data-Sense.md

[davclark]

I'm happy to try filling it out from the above / with notes from you all. Or, you can edit directly - it's markdown, and will be rendered by Jekyll / GitHub pages.

[dnafus]

Dav, So it looks like some of the earlier links for this have shifted, but I see from here http://hackingmeasurement.berkeley.edu/syllabus/2015/09/22/Intel-Data-Sense.html that we ought to get our act together in terms of how the session should go.

I’m thinking the students should be given access beforehand—how many students are there likely to be? I can generate a code if I have a ballpark figure.

In terms of pre-reading, I’m enclosing an introduction to my book that may or may not be appropriate for this kind of class. I worry that it might be too dense/too social scientific for them, but it’s on offer. A much lighter piece that focuses on data sense specifically is this https://www.epicpeople.org/data-data-everywhere/ which was written for my professional association.

What kinds of things would you like to do in the context of the session? Will students have ongoing projects we can talk about? Should we drum up some data set they are interested in working with and see what we see? Anything we should know about what Joe is going to do before us?

If all this really just sounds like a half hour phone call, just let me know…

Dawn

From: Dav Clark [mailto:notifications@github.com] Sent: Thursday, August 20, 2015 12:05 PM To: hackingmeasurement/hackingmeasurement.github.io Cc: Nafus, Dawn Subject: Re: [hackingmeasurement.github.io] Data management platforms (#22)

I'm happy to try filling it out from the above / with notes from you all. Or, you can edit directly - it's markdown, and will be rendered by Jekyll / GitHub pages.

— Reply to this email directly or view it on GitHubhttps://github.com/hackingmeasurement/hackingmeasurement.github.io/issues/22#issuecomment-133122205.

[davclark]

First, I'm realizing that you attached a document that I think GitHub stripped. Can you either email that to me directly, or just drag n' drop it into the issue on the GitHub issue page?

I expect <20 students. 25 codes seems very safe / conservative. I'm hoping to entice Hélène Mialet to come. She's been working with Type-I diabetics in particular, and I think your platform / approach would be a nice complement to her digging into disease-specific communities of quantification.

Project selection is going slow. However, there will probably be something built around NeurotechX. I've been talking with @javirosa about doing some stuff with human movement. Hopefully we'll have a bit of a kit folks can throw in their backpacks to track movement and temperature. It'd be fun to start thinking about sticking movement sensors on parts of themselves. Another participant is very interested in transit movement.

[dnafus]

Excellent.

So direct the students here: https://makesenseofdata.com/event/invites with the code “hackingmeasurement”.

Will send paper direct to your email. Would be fabulous to have Hélène come!

From: Dav Clark [mailto:notifications@github.com] Sent: Monday, September 14, 2015 2:28 PM To: BIDS-collaborative/hackingmeasurement.github.io Cc: Nafus, Dawn Subject: Re: [hackingmeasurement.github.io] Data management platforms (#22)

First, I'm realizing that you attached a document that I think GitHub stripped. Can you either email that to me directly, or just drag n' drop it into the issue on the GitHub issue page?

I expect <20 students. 25 codes seems very safe / conservative. I'm hoping to entice Hélène Mialet to come. She's been working with Type-I diabetics in particular, and I think your platform / approach would be a nice complement to her digging into disease-specific communities of quantification.

Project selection is going slow. However, there will probably be something built around NeurotechX. I've been talking with @javirosahttps://github.com/javirosa about doing some stuff with human movement. Hopefully we'll have a bit of a kit folks can throw in their backpacks to track movement and temperature. It'd be fun to start thinking about sticking movement sensors on parts of themselves. Another participant is very interested in transit movement.

— Reply to this email directly or view it on GitHubhttps://github.com/BIDS-collaborative/hackingmeasurement.github.io/issues/22#issuecomment-140210645.

[dnafus]

Okay, so it sounds like they are working together on one thing, as opposed to working independently? I do agree that starting with yourself is really the best way to learn about what data can/cannot do. Maybe our role can be one of just denaturalizing the tendency to assume "data" is just spat out the other end of the sensor, and give them some options with respect to cooking it/communicating it. Would that help or is this terrain they already know well? We have lots of time-- my current thinking is to do half hour or so talk on this topic, a q&a, then flip it over to them and have them talk about their project/brainstorm strategies for handling or representing data. If you have an example of data types they are likely to work with, I could throw it on Data Sense and we can explore it together to get a feel for what they might encounter in their projects.

Something like that?

[davclark]

Based on your putting the code above, I wonder if you remember that this issue is a public record on GitHub? But there's probably sufficient security through obscurity.

In the class, we've all started a basic Arduino project as a training example. However, teams are finally forming and students will be going in different directions, ranging from air quality and seismic sensing to health data and so on. I have a big interest in tutoring / education, but no bites on that yet sadly!

By the time we get to your presentation, we will probably have just gotten started with folks collecting acceleration and temperature (and maybe other) data from a little self-contained arduino platform in their backpack. The idea is we should be able to get a better sense of indoor / outdoor from this.

So, probably some movement data from accelerometers would be awesome, and if you wanted to add something else, cardiogram would also be great. We'll be using adhesive electrode pads, so should be able to get reasonable HRV. But I'm not sure until I see the data off the kit... pulse oximeter or similar would also be fine.

I'm also personally very interested in breath, and the sensor from spire.io.

[dnafus]

I had forgotten that—obscurity is probably enough… I’ll keep an eye on the site traffic with that in mind—thanks for letting me know!

I looked at spire for another project, but there’s no export yet—but I’ve been wanting to try it too!

I’ll plan on having some data in-hand for students to play with, along the lines you mention (I have some AQ data on hand, too), and maybe we can try fiddling with getting the data out of what they’ve started to collect, and onto Data Sense to see what we can see?

From: Dav Clark [mailto:notifications@github.com] Sent: Wednesday, September 16, 2015 11:56 AM To: BIDS-collaborative/hackingmeasurement.github.io Cc: Nafus, Dawn Subject: Re: [hackingmeasurement.github.io] Data management platforms (#22)

Based on your putting the code above, I wonder if you remember that this issue is a public record on GitHub? But there's probably sufficient security through obscurity.

In the class, we've all started a basic Arduino project as a training example. However, teams are finally forming and students will be going in different directions, ranging from air quality and seismic sensing to health data and so on. I have a big interest in tutoring / education, but no bites on that yet sadly!

By the time we get to your presentation, we will probably have just gotten started with folks collecting acceleration and temperature (and maybe other) data from a little self-contained arduino platform in their backpack. The idea is we should be able to get a better sense of indoor / outdoor from this.

So, probably some movement data from accelerometers would be awesome, and if you wanted to add something else, cardiogram would also be great. We'll be using adhesive electrode pads, so should be able to get reasonable HRV. But I'm not sure until I see the data off the kit... pulse oximeter or similar would also be fine.

I'm also personally very interested in breath, and the sensor from spire.io.

— Reply to this email directly or view it on GitHubhttps://github.com/BIDS-collaborative/hackingmeasurement.github.io/issues/22#issuecomment-140842492.

[davclark]

Sounds great! When I reach out to Neema, I'll mention your interest in getting data from spire as well.