How to create a scatter plot with yticklabels in a custom order

Question

can I do this and produce a scatter?

y-axis tick/label order → [0, 7, 2, 9, 4, 11, 6, 1, 8, 3, 10, 5]
            
x = ('a', 'b', 'c', 'd', 'e')

y = [[1,5,2], [10,5,11,7], [9], [], [ 7, 2, 9, 4, 11, 6, 1, 8, 3, 10, 5]]

The idea is to map the x-values against the y-values. e.g. the first in the list of data ('a') was, say, hitting 1, 5 and 2 on a target, the second ('b') hit scores 10, 5, 11, 7. I need to keep the Y-axis as labels in the order shown.

Answer 1

The implementation consists of getting the y-axis ticklabels in the order defined by y_ticks, which requires changing the yticklabels to something that does not match the default ytick coordinates.

import matplotlib.pyplot as plt

# Define the data we'll work with. 
x_labels = ("a", "b", "c", "d", "e")
x_ticks = range(len(x_labels))
y_ticks = (0, 7, 2, 9, 4, 11, 6, 1, 8, 3, 10, 5)
# y values to be plotted
y_lists = ([1, 5 , 2], [10, 5, 11, 7], [9], [], [7, 2, 9, 4, 11, 6, 1, 8, 3, 10, 5])

# Define the figure and ax.
fig, ax = plt.subplots()
fig.show()

# Disable y-autoscaling since we are defining the ticks and ticklabels ourselves.
ax.autoscale(False)

# We can use a bit of custom padding.
e = 0.05

xmin, xmax = min(x_ticks), max(x_ticks)
dx = xmax - xmin
ax.set_xlim(xmin - dx*e, xmax + dx*e)

ymin, ymax = min(y_ticks), max(y_ticks)
dy = ymax - ymin
ax.set_ylim(ymin - dy*e, ymax + dy*e)

# Set the sorted y-ticks.
ax.set_yticks(sorted(y_ticks))
ax.set_yticklabels(y_ticks)

# Set the x-ticks.
ax.set_xticks(x_ticks)
ax.set_xticklabels(x_labels)  # `ax.set_xticklabels("abcde")` would work too.

for x, y_list in zip(x_ticks, y_lists):
    # We have a single x value for each letter, but we need the x-list to be as long y-list in order to make a scatter.
    x_list = [x]*len(y_list) 
    # Notice the use of `.index` here to accommodate the ticks not being ordered.
    true_y_list = [y_ticks.index(y) for y in y_list] 
    ax.scatter(x_list, true_y_list)

For a variant where a dict is used to hold the data and automatically name the x-ticks after its keys, we can replace the first paragraph by this:

# Define the data we'll work with.
data = {"a": [1, 5 , 2],
        "b": [10, 5, 11, 7],
        "c": [9],
        "d": [],
        "e": [7, 2, 9, 4, 11, 6, 1, 8, 3, 10, 5]}

x_labels = data.keys()
x_ticks = range(len(x_labels))
y_ticks = (0, 7, 2, 9, 4, 11, 6, 1, 8, 3, 10, 5)
y_lists = data.values()