…ead of specifying on the command line when running the qa_scene.py script. Also added colored print statements to make the output easier to read. Also added a qa_scene_utils.py file to separate the important processing code from helper functions

…Application so that we can test how fast the rendering is

…abstracted some big chunks of scene/stage processing code into functions, and created a new qa scene json config file

…ate unique filenames for output files

…ofile_scene() from inside the iteratively_test_all_scenes() function. Added ability to specify initial transform of agent. Started writing code to generate a default scene if none exists in the dataset config file

…u specify in the config file. You can press 'c' to watch the collision test occur, or if you don't run the app with the viewer on, it will run all the tests specified in the config file you want to test. Refactoring some of the functions in qa_scenes.py, and abstracting some code snippets into separate methods bc some of the functions were getting long and difficult to read. Also added code to time how long the collision test takes at each position

…er times to the comprehensive json

…over a given number of seconds. It saves the results in the comprehensive json file

…object sleep test

…aries for each scene

…till work when some of the tests aren't run, e.g. if we are running the render test, but not the collision nor sleep tests

…Now runs on both stages and scenes

…nd path before it

…some config files and updated the old ones

…better crash tracking

…pv0.2.1

…ig options for reduced metric output. Minor cleanup.