trajectorytools 0.4

A tool to study 2D trajectories

trajectorytools is a library with some utils to study and plot 2D trajectories.

Installation

From PyPI

pip install trajectorytools

From Source

pip install git+https://gitlab.com/polavieja_lab/trajectorytools

Example

import numpy as np
from matplotlib.cm import ScalarMappable
from matplotlib.colors import Normalize

import trajectorytools as tt
import trajectorytools.animation as ttanimation
import trajectorytools.socialcontext as ttsocial
from trajectorytools.constants import test_raw_trajectories_path

# Loading test trajectories as a numpy array of locations
test_trajectories = np.load(test_raw_trajectories_path, allow_pickle=True)

# We will process the numpy array, interpolate nans and smooth it.
# To do this, we will use the Trajectories API
traj = tt.Trajectories.from_positions(test_trajectories, smooth_params={"sigma": 1})

# We assume a circular arena and populate center and radius keys
center, radius = traj.estimate_center_and_radius_from_locations()

# We center trajectories around the estimated center
traj.origin_to(center)

# We will normalise the location by the radius:
traj.new_length_unit(radius)

# We will change the time units to seconds. The video was recorded at 32
# fps, so we do:
traj.new_time_unit(32, "second")

# Now we can find the smoothed trajectories, velocities and accelerations
# in traj.s, traj.v and traj.a
# We can use, for instance, the positions in traj.s and find the border of
# the group:
in_border = ttsocial.in_alpha_border(traj.s, alpha=5)

# Animation showing the fish on the border
colornorm = Normalize(vmin=0, vmax=3, clip=True)
mapper = ScalarMappable(norm=colornorm, cmap="RdBu")
color = mapper.to_rgba(in_border)

anim1 = ttanimation.scatter_vectors(traj.s, velocities=traj.v, k=0.3)
anim2 = ttanimation.scatter_ellipses_color(traj.s, traj.v, color)
anim = anim1 + anim2

anim.prepare()
anim.show()

In the directory examples, you can find some more example scripts. Scripts use some example trajectories, which can be found in data. All example trajectories were obtained using idtracker.ai on videos recorded in de Polavieja Lab (Champalimaud Research, Lisbon)

— NOTE

Note that, when using constructors like from_idtrackerai and from_positions, we need to calculate velocity and accelerations from positions. As a result, the traj object has 2 frames less than the original positions array. By default, the missing frames correspond to the first and last frames of the video. If you used the option “only_past”:True in smooth_params, the missing frames correspond to the first two frames of the video.

—

Project maintainers

Jordi Torrents (2024-) Dean Rance (2021-2021) Francisco J.H. Heras (2017-2023) Francisco Romero Ferrero (2017-2023)

Contribute

We welcome contributions. The preferred way to report problems is by creating an issue. The best way to propose changes in the code is to create a pull request. Please, check our contribution guidelines and our code of conduct.

License

This project is licensed under the terms of the GNU General Public License v3.0 (See COPYING). This means that you may copy, distribute and modify the software as long as you track changes/dates in source files. However, any modifications to GPL-licensed code must also be made available under the GPL along with build & install instructions.

If you use this work in an academic context and you want to acknowledge us, please cite some of the relevant papers:

Romero-Ferrero, F., Bergomi, M. G., Hinz, R. C., Heras, F. J., & de Polavieja, G. G. (2019). idtracker.ai: tracking all individuals in small or large collectives of unmarked animals. Nature methods, 1

Heras, F. J., Romero-Ferrero, F., Hinz, R. C., & de Polavieja, G. G. (2019). Deep attention networks reveal the rules of collective motion in zebrafish. PLoS computational biology, 15(9), e1007354.