Scatter Plots

Part of the Data and Plots section of Coddy's Machine Learning journey. Lesson 21 of 60.

Ten houseplants: hours of sun a day, and how much each grew. Joined by a line, they make a tangle. As dots, a pattern appears:

Two charts of the same ten plants, sun across and growth up. Left, plt.plot joins them into a zigzag tangle. Right, plt.scatter shows dots rising to the right

plt.scatter(x, y) draws one dot per pair, with no lines:

sun = [2, 6, 3, 8, 5, 1, 7, 4, 6, 3]
growth = [4, 10, 7, 12, 6, 2, 9, 8, 7, 3]
plt.scatter(sun, growth)
plt.xlabel("Sun (hours a day)")
plt.ylabel("Growth (cm)")
plt.show()

More sun, more growth. A scatter plot shows how two measurements of separate samples relate: the first look before a model predicts one from the other.

Values in order, like the days of a week, still get plt.plot.

Challenge

Medium

An angler logged each fish in catch order: one row per fish, length (cm) then weight (g). The starter joins them with plt.plot, so the line zigzags and hides the pattern.

Draw every fish as its own dot: length across, weight up.
Leave no line at all.
Keep the title, the labels and plt.show().

Yours should look like this one, from another trip:

A scatter plot titled Fish Catch: nine dots rising from 18 cm and 90 g to 42 cm and 980 g, with no line

Your program prints nothing. Coddy describes your chart in the output.

Try it yourself

import numpy as np
import matplotlib.pyplot as plt

# Each row is one fish, in the order it was caught: length (cm), weight (g)
fish = np.array([[22, 180], [30, 390], [25, 250], [35, 620],
                 [28, 330], [40, 870], [19, 120], [33, 510]])

# Fix the drawing below
plt.plot(fish[:, 0], fish[:, 1])

# Don't change the lines below
plt.title("Fish Catch")
plt.xlabel("Length (cm)")
plt.ylabel("Weight (g)")
plt.show()