Modifications to master_openworm.py

[slarson]

initial version

• [x] Step 1: Call sibernetic_c302 script after setting environment variables to run Sibernetic.
  • [x] Run c302 with the NIPS version of the configuration.
  • [x] Modify sibernetic_c302 to produce the graph images before it starts running Sibernetic
  • [x] Images will be produced -- save images to a known "output" directory
  • [x] As sibernetic is running and outputting file with positions of all particles as result, this file should also be in the "output" directory
  • [x] Default sibernetic run time should be 4 minutes
• [x] Step 2: generate movie from output file
  • [ ] Also generate curvature and movement graph plot images from output file
  • [x] Save all contents to the known "output" directory

The user will see stdout statements while system is running letting them know what is happening. As steps complete, they will get the graphs from the nervous system simulation and then they will get a movie + movement plots of the sibernetic movement.

[slarson]

For getting Sibernetic to work via docker, note the comments in this issue: https://github.com/openworm/sibernetic/issues/90

[slarson]

May want to reference the documentation on this old docker attempt: https://hub.docker.com/r/jamiemori/openworm_test/

[slarson]

Generating movie from output file looks like the next, complicated step. Looking at this and a way of adapting it to work here. xvfb-run and ffmpeg are almost certainly the right tools to use here. The path would be to first run Sibernetic headlessly for speed, then replay the position file via OpenGL into a virtual framebuffer via xvfb-run, and connect ffmpeg to the framebuffer to save the frames as .mp4. Devil is in the details :)

[slarson]

Incorporated into 0.7 release: https://github.com/openworm/OpenWorm/releases/tag/0.7