Density scatterplots using Turbo colormap from google

density_scatter.py

from matplotlib import pyplot as plt
import numpy as np
from scipy.interpolate import interpn

# Density scatterplots, based on https://stackoverflow.com/a/53865762
def density_scatterplot(x, y, bins=100, cmap='turbo', drop_na=True, ax=None, **kwargs):
    if ax is None:
        ax = plt.gca()
    x = np.asarray(x)
    y = np.asarray(y)

    if drop_na:
        na_mask = np.isnan(x) | np.isnan(y)
        x = x[~na_mask]
        y = y[~na_mask]

    # Based on https://stackoverflow.com/a/53865762
    data , x_e, y_e = np.histogram2d(x, y, bins = bins, density = True )
    z = interpn(
        ( 0.5*(x_e[1:] + x_e[:-1]),
         0.5*(y_e[1:]+y_e[:-1]) ),
        data,
        np.vstack([x,y]).T,
        method = "splinef2d",
        bounds_error = False
    )

    # To be sure to plot all data
    z[np.where(np.isnan(z))] = 0.0

    # Sort the points by density, so that the densest points are plotted last
    idx = z.argsort()
    x, y, z = x[idx], y[idx], z[idx]

    return ax.scatter(x, y, c=z, cmap=cmap, **kwargs)