Notebook Entry | moorepants

AIO for March 2, 2015

Last week's accomplishments

Yeadon Paper

[5 hr, Done 4.7 hr] Finish and submit the second revision by March 1.
- Second revision is here: https://github.com/chrisdembia/python-yeadon-paper

Perturbed Data Paper

Job Search

[5 hr, Done 3.4 hr] Finish teaching statement draft and ask for reviews from Rebecca, Sarah, and Cass.
[1 hr, Done 1.5 hr] Request reference letters for lecturer position.
[1 hr, Done 1.5 hr] Second interview with robot/machine learning company.
[0.5 hr, Done] Respond to Sherm and setup a day to do the coding test for the Opensim job.
[2.5 hr, Done 3 hr] Meet with Prof. Kong at UCD about post doc.
[9.7 hr] Interviewed with sports biomechanics startup.
[30 min] Turned down the opensim job.

Perturbed Data Paper

[5 hr] Fix as many issues as possible in the alloted time. (20 days left till due date).
- See below. Didn't make time.

Quiet Standing Identification Paper

[20 hr, Failed] Finish first draft by February 28.
- I went for some spontaneuous itnerviews in the bay area instead of meeting this deadline. I did get some work done but managed to booger the direct collocation solutions by trying a "better" input perturbation. I think this threw off the scaling. I'm going to generalize the scaling so hopefully it works. I was trying to have a sum of sines with a human scale bandwidth with decreasing amplitude vs frequency for a realistic "random" input. Got some tips from my old PI on this.
- I also updated some of the paper, but the results are still lackluster.

Job Search

Perturbed Data Paper

[5 hr] Fix as many issues as possible in the alloted time. (20 days left till due date).

Quiet Standing Identification Paper

[20 hr] Finish first draft by end of week.
- Introduce optimal random input.
- Generalize scaling.
- Create plot to show the optimal step size in direct collocation.
- Rerun the 10,000 identifications with the optimal step size.