Supporting progressive visualization with degree-of-interest functions.
Prerequisites
- requires nodeJS version 16 or later, with npm version 7 or later
- requires python 3.8 (NOTE: due to a dependency on the
pyscagnosticspackage, later versions than 3.8 are currently NOT supported), with pip version 22.
Getting Started
To install ProInterest, you need to install both the frontend and backend components, located in their own subdirectories in this repository.
Installing the Frontend
To install the frontend component, open a terminal in the client/ directory and download the dependencies via npm:
npm install
Installing the Backend
To install the backend component, open a terminal in the server/ directory and download the dependencies via pip:
python -m pip install -r requirements.txt
Downloading the Data
ProInterest uses the NYC taxi dataset, which can be downloaded from NYC OpenData. After downloading, place the data in the server/data/ directory, with the file name nyc_taxis.unshuffled_full.csv.gz. Generate a .parquet file by running the csv_to_parquet.py script in the same directory, passing the csv.gz file as a parameter.
Getting Started
With both frontend and backend installed, you can now launch ProInterest in your browser:
First, launch the server from your terminal inside the server/ directory via
python server.py
and afterwards launch the frontend via
npm run dev
After a few seconds, you can navigate your browser to https://localhost:3000.
Running the Benchmarks
The benchmarks we report on in our paper are located in the server/benchmarks/ directory.
The test_cases.json file defines presets you can use to replicate our results, by running
python benchmarks.py <index in test_cases.json> composite
Afterwards, you can visualize the results inside the analysis.ipynb notebook.