Collecting Data at E’s Bar

In ITP, Quantified Self

For this week, we did an initial data collection round to see if we could see trends in the data. Matt and I decided to narrow our focus to trying to quantify the ambience in a bar, as we think it’s super important to give ourselves more constraints in an attempt to make the data more meaningful. Our goal now is to look at the light, sound and temperature in various bars and match that with¬†qualitative observations we make every 20 minutes while recording. Here is a plot of the three sensors over and hour long period from 5:49 to 6:49 at a bar on the Upper West Side on Monday evening.



As you can see, there are a couple clear issues right out the gate. We need to make the audio and light readings more proportional to make their relationship more apparent. We noticed that there is a slight correlation between the light readings dipping and the sound readings peaking but we need to confirm this. Currently, we are mapping the light readings from 0-1023 to 0-255 and I think simply removing that should do the trick. Here are the qualitative assessments I made below. For the final presentation, we’d like to find a way to visually include the qualitative assessments into the graph.


Observations at 5:49

  • Light volume of people, roughly 15-18 people in bar
  • Good after work vibes, some excitement
  • Mellow music

Observations at 6:09

  • Slightly more people, slightly more animated (probably about 5 more people in bar)
  • evening starting to set in
  • animated but laid back

Observations at 6:29

  • Same amount of people
  • Slightly louder it seems, maybe because of more alcohol
  • Mostly constant from last reading

Observation at 6:49

  • Sun setting, candles came out
  • Energy and noise stable
  • Still mellow laid back week night


Submit a comment