Skip to content

Note

Click here to download the full example code

Spatial Scatter Bias Plot (Alternative)

What it's for: This plot (intended as an alternative to Spatial Bias Scatter) visualizes model bias at geographic locations.

When to use: Use this to identify regional patterns in model errors, specifically comparing a reference dataset (observations) with a comparison dataset (model output) across a geographic domain.

How to read: * Markers: Represent data points at specific Latitude/Longitude coordinates. * Color: Represents the bias (Comparison - Reference). * Interpretation: Clusters of similar colors indicate regional systematic biases in the comparison dataset.

Note: This specific example is currently a placeholder for an alternative implementation.

import numpy as np
import pandas as pd

# from monet_plots.plots.sp_scatter_bias import SpScatterBiasPlot

# 1. Prepare sample data
np.random.seed(42)  # for reproducibility
n_points = 500

# Simulate random latitude and longitude points
lat = np.random.uniform(20, 50, n_points)
lon = np.random.uniform(-120, -70, n_points)

# Simulate reference and comparison values
reference_values = 10 + 5 * np.random.rand(n_points)
# Introduce a spatial bias: higher values in the west, lower in the east
comparison_values = reference_values + (lon / 100 + np.random.normal(0, 0.5, n_points))

df = pd.DataFrame(
    {
        "latitude": lat,
        "longitude": lon,
        "reference_value": reference_values,
        "comparison_value": comparison_values,
    }
)

# 2. Initialize and create the plot
# Example skipped: SpScatterBiasPlot module not found in this version.

Total running time of the script: ( 0 minutes 0.001 seconds)

Download Python source code: plot_sp_scatter_bias.py

Download Jupyter notebook: plot_sp_scatter_bias.ipynb

Gallery generated by mkdocs-gallery