Getting started#

Hi there and welcome to the getting started guide of TASI 👋😀 This page shall help to get started right away with instructions on how to install TASI and a description of the main TASI concepts. You may visit the following links one after another. Note that the sections may contain links to external pages. It is recommended to visit those links before continuing. If you have any questions regarding TASI, feel free raise an issue. Happy reading 😁

Installation#

An always up-2-date version of TASI is available via the Pypi registry. So, to get started with TASI, its just a matter of installing tasi with

pip install tasi

Basic concepts#

Similar to pandas and geopandas in TASI, we represent traffic data using a tabular-styled format. To achive this, TASI extends pandas and geopandas to provide functionality for traffic data analysis. The top-level model is the Dataset as illustrated in Fig. 1. Every row in a Dataset contains information for a specific point in time as attributes. That is, an attribute can be any relevant traffic information, e.g., the position of a traffic participant or the rain intensity.

TASI dataset model — Fig. 1 The top level model for traffic data representation is the `Dataset`.#

Most of the time, we will work with information about the traffic participants that are represented in the Dataset. To make live a little bit easier, TASI uses a hierarchical view on trajectory data based on the Dataset as illustrated in Fig. 2. Since a Dataset typically contains information about multiple traffic participants, each having a trajectory, the Dataset is a container for multiple Trajectory. This is denoted with the different id values and color encoded rows. Furthermore, since a single row is a representation of a traffic participant’s state in a specific point in time, the Pose

representation of a traffic participant’s state for a specific point in time is with the class.

TASI trajectory and pose model — Fig. 2 Hierarchical view on trajectory data.#

Since TASI uses Pandas under the hood, head over to the official pandas User Guide. Afterwards, you may head over to the User Guide for some use-case related examples using TASI. If you want to contribute to TASI, visit Development Documentation for further instructions.