This new accelerometer recording system is quite interesting. I had to tweak the protocol a bit, because A) the phone would go to sleep during the recording, which stops the flow of data B) I had to figure out how to position the phone near my center of gravity, to leave out extraneous motions like the swinging of my jacket or pants pocket and C) how to do all this without looking like a total retard. Anyway, Valentine’s Day saw the best recording yet. (I didn’t bother processing the earlier, incomplete data sets.) Here are single-dimension graphs, showing the main axes. There are XYZ graphs for the morning and the evening:
Notice the smoother areas where I’m riding the subway. When I’m walking, my body is undergoing a constant push-and-pull against gravity, basically bouncing up then lurching down with each step, which creates the spiky waveform. Also, notice that the Y-axis acceleration tends to drift to -10 m/s^2. That’s because the phone was upside-down, and the acceleration due to Earth’s gravity is 9.8 m/s^2. Excuse me while I geek out a little on that.
I’m working on an easier way to produce 3D graphs of the data in scatterplot, so I’ll post those tomorrow.