Density profiles

[chraibi]

Density profiles with Voronoi are great and precise.

However, sometimes you may have a big geometry or a lot of people moving around.
This means the profile calculations can be very slow.

At that scale, it might be enough to have an imprecise density profile, but fast.

Therefore, I suggest introducing new density profiles:

1. Classical method
2. Gaussian method
   (published in this paper)

You can try both methods here.

Here some equations:

[schroedtert]

The suggested functions to compute the density profiles have been added to PedPy. With the next release (1.1) they will available via PyPi, until then you can check them by installing the latest (unstable) version of PedPy via pip install git+https://github.com/PedestrianDynamics/PedPy.git.

To use them you can do the following:

from pedpy import compute_density_profile, DensityMethod

grid_size = 0.25

classic_density_profile = compute_density_profile(
    data=traj.data,
    walkable_area=walkable_area,
    grid_size=grid_size,
    density_method=DensityMethod.CLASSIC,
)

gaussian_density_profile = compute_density_profile(
    data=traj.data,
    walkable_area=walkable_area,
    grid_size=grid_size,
    density_method=DensityMethod.GAUSSIAN,
    gaussian_width=0.5,  
)

Thank you, for the valuable input!