Tracking in nD: Linking

Using whatever trick you have in your notebook, finally you get some coordinates. Then you want the trajectories of the agents/fish/particles. How would you convert a collection of positions into trajectories?

Very briefly, I would write a script like below:

import fish_track as ft

linker = ft.ActiveLinker(linker_range)  # alternatively, use ft.TrackpyLinker
trajectories = linker.link(positions)
trajectories_longer = ft.relink(trajectories, dx, dt, blur)

In the end, we get trajectories_longer which are long trajectories. The above code works very nicely in any dimension. (If you have experimental data for particles in 4 or higher dimension, please send me an email so I can worsihp your techniques.)

But what is happening? What are the meanings of those variables? What is the business happening inside the code?

The Paradox of Linking

under construction …

The Colloidal Linker

under construction …

The Active Linker

under construction …

Making Trajectories Longer

under construction …