Andrews curve is a method that's used to visualize multidimensional data. It does this by mapping each observation onto a function. Here, each color that's used represents a class and we can easily note that the lines that represent samples from the same class have similar curves. This curve is very useful in analyzing time series and signal data.

Basically, each data point is sent through a Fourier transform according to the Fourier function. Each line in the following chart represents a separate data point. It can be plotted using the snippet below.

andrew = pd.plotting.andrews_curves(mtcars,'Category') 

The following is the graph: