Show statistics and make plots extracted from jpscore-simulations and experimental data.
This app can be used online:
or locally by running
streamlit run app.py
To install the requirements use:
pip install -r requirements.txt
To draw geometries on trajectory-plots, try to connect all lines, such that they form a closed polygon. This is only necessary if you want to visualize the geometry with jpsvis.
You can download your drawing as an xml file according to jupedsim's format
It is possible to draw:
- Lines
- Rectangles
Lines and rectangle can be rotated and scaled.
To try out: https://go.fzj.de/geometry
- Demo: https://youtu.be/4xTqSbllCwg
- Talk in the deRSE23 from some scientific motivation of this project is available here.
NOTE: The script tries to exctract the unit of the data from the trajectory file. A good idea might be to convert the trajetories before using with this script
or add a header with the unit information. For example a line starting with
#unit: cm
The density profile uses the speed of the pedestrians calculated by the Weidmann diagram.
Therefore, for jpscore-simulations, use the optional_output
-option in the inifile:
<file location="bottleneck_traj.txt" />
<optional_output speed="TRUE"/>
</trajectories>
NT-curves and flow at lines (transitions
or area_L)
are calculates be testing for every agent, if croses the line.
Therefore, define in the geometry transitions, such that the pedestrian fully pass them.