urban-toolkit / utk

A Grammar-based Framework for Urban Visual Analytics
http://urbantk.org
MIT License
68 stars 6 forks source link
urban-analytics urban-data visual-analytics visualization

PyPI Slack

UTK logo

The Urban Toolkit (UTK)

The Urban Toolkit is a flexible and extensible visualization framework that enables the easy authoring of web-based visualizations through a new high-level grammar specifically built with common urban use cases in mind.

For a quick getting starter document and tutorials, visit: urbantk.org

UTK was first presented in the paper:
The Urban Toolkit: A Grammar-based Framework for Urban Visual Analytics
Gustavo Moreira, Maryam Hosseini, Md Nafiul Alam Nipu, Marcos Lage, Nivan Ferreira, Fabio Miranda
IEEE Transactions on Visualization and Computer Graphics, 2024

@ARTICLE{utk_2024,
  author={Moreira, Gustavo and Hosseini, Maryam and Alam Nipu, Md Nafiul and Lage, Marcos and Ferreira, Nivan and Miranda, Fabio},
  journal={IEEE Transactions on Visualization and Computer Graphics}, 
  title={{The Urban Toolkit}: A Grammar-Based Framework for Urban Visual Analytics}, 
  year={2024},
  volume={30},
  number={1},
  pages={1402-1412},
  doi={10.1109/TVCG.2023.3326598}
}

UTK cases

Table of contents

  1. Features
  2. Installation and quick start
    1. UTK backend
    2. UTK frontend
  3. Tutorials
  4. Development
    1. Slack channel
  5. Other resources
  6. Team

Features

UTK has been tested on Linux (Ubuntu 23.04), Windows 10 & 11, and MacOS 13.

Installation and quick start

UTK leverages several spatial packages, such as Geopandas, OSMnx, Osmium, Shapely. To facilite the installation of UTK, we have made it available through pip, only requiring the following commands in a terminal / command prompt:

pip install utk

UTK requires Python 3.9, 3.10 and 3.11 (there is an issue with 3.12 that is on our TODO list to solve). If you are having problems installing UTK in Mac OSX because of Osmium, make sure you have CMake installed as well (e.g., through conda or Homebrew).

A detailed description of UTK's capabilities can be found in our paper, but generally speaking UTK is divided into two components: a backend component, accessible through UTK's Python library, and a frontend component, accessible through a web interface.

UTK backend

UTK's backend is available through our Python library. For example, using a Jupyter Notebook, it can be imported with:

import utk

To download data for Manhattan, NY, you only need to then:

uc = utk.OSM.load('Manhattan, NY', layers=['surface', 'parks', 'water', 'roads'])
uc.save('./manhattan')

This will create a new folder (manhattan) with the downloaded and parsed OSM data. On top of that, UTK also offers functionalities to load data from shapefiles (utk.physical_from_shapefile), csv files (utk.thematic_from_csv), dataframes (utk.thematic_from_df), and also accumulate sunlight access values (utk.data.shadow). A detailed description of UTK's Python API can be found here.

UTK frontend

UTK's frontend is available through the utk command. After the pip installation is complete, you can check a toy example with the following commands:

utk example

UTK's toy example will then be accessible through localhost:5001. After accessing it using a browser, you should see a grammar-defined visualization showing sunlight access per building:

UTK example

Beyong the simple example, you can also use the utk command to start and stop UTK's server:

utk start
utk stop

The utk command takes the following arguments:

usage: utk [-h] [-d [DATA]] [-b [BUNDLE]] [-g [GRAMMAR]] [-a [ADDRESS]]
           [-p PORT]
           {start,list,stop,example}

The Urban Toolkit

positional arguments:
  {start,list,stop,example}
                        Start, list or stop utk servers, or start server with
                        a simple example.

optional arguments:
  -h, --help            show this help message and exit
  -d [DATA], --data [DATA]
                        Path to data folder.
  -b [BUNDLE], --bundle [BUNDLE]
                        Path to app bundle (defaults to installed utk bundle).
  -g [GRAMMAR], --grammar [GRAMMAR]
                        Path to grammar JSON file, if different from
                        [DATA]/grammar.json (default: [DATA]/grammar.json).
  -a [ADDRESS], --address [ADDRESS]
                        Server address (default: localhost).
  -p PORT, --port PORT  Server port (default: 5001).
  -w, --watch           Watch folders, and re-build if there are changes.

Even though we offer support for a variety of arguments, most users will simply need to run the following to use data stored in a folder called ./data/:

utk start --data ./data

After starting UTK's server and opening localhost:5001 on a browser, you will see UTK's main interface, composed of a grammar editor (left) and map viewer (right). Adding new elements to the grammar specification on the right (e.g., new plots, new data) will automatically update the map viewer:

UTK example

Simulations

Currently supported simulations:

Tutorials

UTK tutorials

Step-by-step tutorials are available on our website. These tutorials highlight how UTK can be used to create sophisticated urban visualizations.

A detailed description of UTK's grammar can be found here.

Development

If you would like to modify UTK's core code, you won't be able to use the utk command (since it points to the utk pip installation). Alternatively, you will have to:

  1. Clone the repository with git clone git@github.com:urban-toolkit/utk.git.
  2. Install Node.js, either using conda, package managers, or pre-built installers.
  3. Build the utk-ts bundle. Inside src/utk-ts, you should run the following in the terminal: npm install && npm run build. After that, a bundle will be created.
  4. Build the utk-frontend bundle. Inside src/utk-frontend, you should run the following in the terminal: npm install && npm run build:web. After that, another bundle will be created.
  5. Run the UTK server. Inside the src, run the following in the terminal: python utk_server.py start --bundle utk-frontend/build/utk-app/, and with the other appropriate arguments (e.g., --data with the path to your data folder).

To automatically build the bundles (steps 2 and 3) when you make changes to the source code, you can run utk_server.py with the --watch argument.

Slack channel

For question, including development ones, join UTK's Slack. Feel free to post questions on the #installation, #quick-start, and#development channels.

Other resources

Team