Libraries
First, we need to import a few libraries:
- matplotlib: for plotting
- seaborn: for making the plots prettier
import matplotlib.pyplot as plt
import seaborn as snsDataset
The dataset that we will use is the iris dataset that we can load using seaborn.
df = sns.load_dataset('iris')Most simple density plot
The following code produces 3 contour plots using seaborn python library. You have to provide 2 numerical variables as input (one for each axis). The function will calculate the kernel density estimate and represent it as a contour plot or density plot.
The aguments of the function kdeplot() are:
x, y: Variables that specify positions on the x and y axes.shade: Controls the presence of a shade.cmap: Colormap.bw_adjust: Bandwidth, smoothing parameter.thresh: number in [0, 1], Lowest iso-proportion level at which to draw a contour line.
Here is how we can create the default 2D density plot:
sns.set_style("white")
sns.kdeplot(x=df.sepal_width, y=df.sepal_length)
plt.show()
Change colors
In order to make the chart more readable, we can add colormap to the plot:
sns.set_style("white")
sns.kdeplot(x=df.sepal_width, y=df.sepal_length, cmap="Reds", fill=True)
plt.show()
Change smoothing parameter
The bw_adjust argument controls the bandwidth of the kernel density estimate. The higher the value, the smoother the plot will be. Here is an example with a bw_adjust value of 0.5:
sns.set_style("white")
sns.kdeplot(x=df.sepal_width, y=df.sepal_length,
cmap="Blues", fill=True, bw_adjust=0.5)
plt.show()
Going further
This post explains how to create a 2D density plot with seaborn.
You might be interested in how to create a hexbin plot and how to create a 2D histogram.
