Bivariate Plot

Overview

The bivariate_plot_df.py allows users to create bivariate scatter plots with color classification based on specified criteria. This visualization technique is useful for exploring the relationship between two continuous variables in a dataset and understanding how they interact. bivariate_plot_raster.py is for raster images similiar to biscale package in R.

Features

• Supports classification of dataframe or raster using quantiles or equal interval.
• Configurable plot aesthetics (labels, titles, sizes).
• Option to overlay shapefiles for additional context on raster bivariate plot.
• Customizable colormaps and figure size.
• Customizable legend generation to help interpret the bivariate classifications.

Installation

To use the function, you need to have Python and the required libraries installed.

• Python 3.x
• numpy
• matplotlib
• rasterio (may require prior gdal installation too)
• geopandas
• os

You can install the necessary packages via pip:

pip install numpy pandas matplotlib rasterio geopandas

Usage

Dataframe bivariate scatter plot

path = r'data\test.csv'
# Call the bivariate classification function using a continuous colormap
bivariate_plot_df(
    path, 'ADF Statistic', 'Keener Z-Statistic',
    style='quantile', 
    n_bins=5, 
    cmap_name='bwr', 
    alpha=0.7, 
    edgecolor='black', 
    plot_kwargs = {},
    legend_kwargs={'legend_position': (1.10, 0.15), 'legend_size': 0.2, 'ticklabelsize': 10}
)

Raster bivariate plot

raster1_path = r'path/to/your/temperature_raster.TIF'
raster2_path = r'path/to/your/precipitation_raster.TIF'
shp_path = r'path/to/your/shapefile.shp'

bivariate_raster_plot(
    raster1_path, 
    raster2_path, 
    shp_path, 
    n_bins=5, 
    style='quantile', 
    cmap_name='coolwarm', 
    legend_kwargs={'ticklabelsize': 10, 'labelsize': 10, 'y_label': 'Precipitation (mm)', 'x_label':'Temperature (°C)'}
)

Contributing

If you'd like to contribute to this project, please fork the repository and submit a pull request with your changes. All contributions are welcome!

License

This project is licensed under the MIT License.