Notebook Entry


AIO for December 9, 2014

Last week's accomplishments

Perturbed Data Paper

  • [18 hr, Closer to done 14 hr] Finish all remaining issues and prepare the draft for submission.
    • Picked a working title
    • Added in journal styles
    • Create participant and trial tables + source code
    • Add syntax highlighting lib
    • Got all exploratory code working with GATK 0.1.2
    • Found to bugs in GATK and made releases 0.1.1 and 0.1.2

Inverted Pendulum Paper

  • [6 hr, Done 6.5 hr] Try out scaling methods and see if convergence improves.
    • I ended up scaling the parameters to be closer to 0.5 and the optimizer now converges nicely, in only ~13 iterations even for long measurement times.

Non-objectives

  • [4.3 hr] Worked with Sabrina on the NAO and came to the conclusion that the methods needed to disable the collision protection were no longer available in the new version 2 of the API.
  • [0.5 hr] Searched for more reviewers for the Yeadon paper.

Last week's issues

I spent a ton of time applying for a job, cover letter, resume updates, website updates, and five interviews. I took Friday off to deal with that.

Yeadon Paper

Gait Control Identification

  • [5 hr] Do some thinking and reasoning about the potential discontinuities in the scheduled gain controller.
    • I didn't spend any time on this.

This week's objectives

Other

  • [3 hr] Fill out termination forms and do whatever I need to do to end my position.
  • [1 hr] Check in with Ton about contract and get that finished up.
  • [1 hr] Clear out books in office to pack this weekend.
  • [2 hr] Finish bicycle paper review for VSD.

Perturbed Data Paper

  • [18 hr] Continue knocking off issues and submit for review.

Inverted Pendulum Paper

  • [5 hr] Write code for the direct identification.
  • [5 hr] Write code for the shooting indirect identification.

Gait Control Identification

  • [5 hr] Do some thinking and reasoning about the potential discontinuities in the scheduled gain controller.