Why does plotting a LinearRegression from sklearn make a crazy graph?

Question

I've seen a couple of posts on how to make a simple graph that looks something like this:

The code for this would be as follows:

# Create linear regression object
regr = linear_model.LinearRegression()

# Train the model using the training sets
regr.fit(X_train, Y_train)

# Plot outputs
plt.plot(X_test, regr.predict(X_test), color='red',linewidth=3)

I've been working on building a similar thing but using the USA_Housing dataset from Kaggle.

My code is as follows:

# plot the graph
plt.scatter(X_test[0], y_test, color = "black")

plt.plot(X_test[0], y_hat)
plt.xticks(())
plt.yticks(())
plt.show()

Which results in the following:

X_test[0] is just one of the independent variables (income).

Shouldn't this just make a simple line through the prediction points?

R-squared is around 91 so the fit is good, but I guess I'm confused as to why it would look like this.

Answer 1

pyplot defaults to '-' for linestyle argument, which means a line that connects through all of the dots, so to produce a scatter plot, whose linestyle are data points instead write

plt.plot(X_test[0], y_hat, '.')

Answer 2

Generally plotting functions connects the points that are consecutive in the array. So, first sort those points and then plot.