#statistics #plot #data #data-science

bin+lib datasaurust

Blazingly fast implementation of the Datasaurus paper

1 unstable release

0.1.0 Apr 10, 2023

#746 in Science

MIT and LGPL-3.0

36KB
612 lines

CI

DatasauRust

Blazingly fast implementation of the Datasaurus paper (500x faster than the original): "Same Stats, Different Graphs: Generating Datasets with Varied Appearance and Identical Statistics through Simulated Annealing" by Justin Matejka and George Fitzmaurice.

Usage

To run with plot -p (using gnuplot):

cargo run --release -- -d data/seed_datasets/Datasaurus_data.csv -p

With pre-defined shape:

cargo run --release -- -p -n 3000000 --decimals 2 --shape cat --allowed-distance 0.1

Starting from Gaussian noise:

cargo run --release -- -p -n 3000000 --decimals 2 --shape cat --allowed-distance 0.1 --gaussian

Create videos

Create video and gif (use --save-plot):

pip install moviepy ffmpeg-python

python scripts/create_video.py logs/cat/ logs/cat.mp4

From one shape to another:

cargo run --release -- -p -n 2000000 --decimals 1 --shape dog --allowed-distance 0.1 --log-interval 10000 -d logs/gaussian_cat/output.csv --save-plots

Note: The original datasets and python code comes from http://www.autodeskresearch.com/papers/samestats

Dependencies

~5–14MB
~189K SLoC