Matplotlib 3.0 Cookbook

There's more...

We have learned how to use twin axes features for sharing the x and y axes individually. Now, let's look at the option of not sharing either of the axes, but using the left and bottom spines for one graph and the top and right spines for the other graph. All four spines use different scales, so there are different ticks and ticklabels.

We will use the familiar Iris dataset for this example. As we know, there are four attributes: petal_width, petal_leangth, sepal_width, and sepal_length. So, as we have seen many times in this book, when we plot a scatter plot with petal_width and petal_length, we see three distinct clusters, but when we plot a scatter plot using sepal_length and sepal_width, we don't observe any such clusters. ...