DrifPredict signature is "DriftPredict(initial_guess, span)" with the comment :
initial_guess : Array of length d. Otherwise assumed to be zero velocity.
span : integer, default 1
Compute velocity field from the most recent span+1 frames.
So I know that I have to give a 2 element array as drift in a 2D plane.... but I don't know which element is "x" (or "y"). So I looked into the source code ( https://github.com/soft-matter/trackpy/blob/master/trackpy/predict.py ) and found that internaly, the drift is a pandas.dataframe converted into a numpy array... but not further.
Maybe someone can tell me the standard order of x and y in trackpy.predict ?
general information:
The python module Trackpy allow to link particles into trajectory. When the particle density is high, linking needs a good initial guess to find the right particle in the next frame. I know my particles have a very deterministic drift along "x" which is near the standard distance between particles, that's why I need to use DriftPredict
A I wrote the question, I dove deeper into the source code. I found
self.pos_columns = kw.get('pos_columns', ['x', 'y'])which is used in calculating the drift.
So I guess the initial_guess parameter should be
np.array(x,y)(or directly(x,y)because it is converted into a numpy.array in the "DriftPredict" object)