Verify behaviour of testSecondaryMatrices

fjandrad commented 6 years ago

Most of the time running the script it seems the workbench matrix is the way to go. Which sounds fishy looking at the results directly on the fitting towards the estimated data.

Possible ways to improve this after digging deeply into the process. are :

[x] Check if putting estimated data in both sensors we get 0 external force
[x] Use estimated data in the ankle sensor to diminish source of error considering if we use estimated data in both sensors we get 0 external force
[x] Calculate wbd offset also on "recalibrated data". See how it changes the performance

fjandrad commented 6 years ago

Verified with commit 8cf1861ee11c7dbdf99ca3f5751e15e19b7853fe and ac4cd60d28445e778c2b98a2de5912cd744edda0.

I will provide plots for showcase improvement later

fjandrad commented 6 years ago

Figures and conclusions using this improvment on the test secondary matrix tests can be found in https://github.com/robotology-playground/insitu-ft-analysis/issues/24#issuecomment-383586172

robotology-playground / insitu-ft-analysis

Verify behaviour of testSecondaryMatrices #29