Question was simple. Get the world position of any item in space throughout the entire timeline, then compare each point to get the frame-by-frame traveled distance.
With pymel, first get the item:
import pymel.all as pm
item = pm.selected()[0]
Get the frame range and turn off cycle check:
start_time = pm.playbackOptions(q=1,animationStartTime=1)
end_time = pm.playbackOptions(q=1,animationEndTime=1)
anim_range = range(int(start_time),int(end_time))
pm.cycleCheck(evaluation = 0)
Then, get its worldMatrix at every frame:
item_values = [pm.getAttr(item.worldMatrix,time=t)[-1][:-1] for t in anim_range]
To figure out the difference between frames, I didn't know how to do, so after some searching, this post had the most concise result:
https://stackoverflow.com/questions/1401712/how-can-the-euclidean-distance-be-calculated-with-numpy
And finally, the code needed to get the differences
per_frame_change = []
for i in range(len(item_values)-1):
per_frame_change.append(numpy.linalg.norm(item_values[i] - item_values[i-1]))
Here's the whole code!
import pymel.all as pm
import numpy
start_time = pm.playbackOptions(q=1,animationStartTime=1)
end_time = pm.playbackOptions(q=1,animationEndTime=1)
anim_range = range(int(start_time),int(end_time))
pm.cycleCheck(evaluation = 0)
item = pm.selected()[0]
item_values = [pm.getAttr(item.worldMatrix,time=t)[-1][:-1] for t in anim_range]
per_frame_change = []
for i in range(len(item_values)-1):
per_frame_change.append(numpy.linalg.norm(item_values[i] - item_values[i-1]))
