You are a Udacity Machine Learning student?
Something like this might be useful for you:
import numpy as np
import matplotlib.pyplot as plt
labels = ['output','varA','varB','varC']
data = np.array([[4,5,6,2,5],
[3,6,4,6,3],
[12,3,5,3,2],
[4,1,1,44, 7]])
colors = ["r", "g", "b", "k"] # make sure you have enough colors to match
# the number of variables
for i in range(1, len(data)):
plt.scatter(data[1], data[i], c=colors[i-1], s=100, alpha=0.7, label=labels[i])
plt.legend()
plt.show()
Hope that helps.
Ronny