Is there a way to label the mean and median in matplotlib boxplot legend?

user_9443671 0
Viewed 3454

I have the following box plot which plots some values with different mean and median values for each box; I am wondering if there is any way to label them so that they appear on the graph legend (because the current box plot plots an orange line for the median and a blue dot for the mean and it is not so clear which is which)? Also is there a way to make one legend for these subplots, instead of having a legend for each one, since they are essentially the same objects just different data?

Here's a code example for one of the subplots, the other subplots are the same but have different data:

fig = plt.figure()
xlim = (4, 24)
ylim = (0, 3700)

plt.subplot(1,5,5)
x_5_diff = {5: [200, 200, 291, 200, 291, 200, 291, 200, 291, 200, 291, 200, 291, 200, 291],
 7: [161, 161, 179, 161, 179, 161, 179, 161, 179, 161, 179, 161, 179, 161, 179],
 9: [205, 205, 109, 205, 109, 205, 109, 205, 109, 205, 109, 205, 109, 205, 109],
 11: [169, 169, 95, 169, 95, 169, 95, 169, 95, 169, 95, 169, 95, 169, 95],
 13: [43, 43, 70, 43, 70, 43, 70, 43, 70, 43, 70, 43, 70, 43, 70],
 15: [33, 33, 39, 33, 39, 33, 39, 33, 39, 33, 39, 33, 39, 33, 39],
 17: [23, 23, 126, 23, 126, 23, 126, 23, 126, 23, 126, 23, 126, 23, 126],
 19: [17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17, 17],
 21: [15, 15, 120, 15, 120, 15, 120, 15, 120, 15, 120, 15, 120, 15, 120],
 23: [63, 63, 25, 63, 25, 63, 25, 63, 25, 63, 25, 63, 25, 63, 25]}

keys = sorted(x_5_diff)
plt.boxplot([x_5_diff[k] for k in keys], positions=keys)  # box-and-whisker plot
plt.hlines(y = 1600, colors= 'r', xmin = 5, xmax = 23, label = "Level 1 Completed")
plt.title("x = 5 enemies")
plt.ylim(0,3700)
plt.plot(keys, [sum(x_5_diff[k]) / len(x_5_diff[k]) for k in keys], '-o')


plt.legend()
plt.show()

Any help would be appreciated.

boxplot matplotlib python
2 Answers

Its a bit late, but try this:

    bp = plt.boxplot([x_5_diff[k] for k in keys], positions=keys)

    # You can access boxplot items using ist dictionary
    plt.legend([bp['medians'][0], bp['means'][0]], ['median', 'mean'])
user_139833890

Store the mean as a separate vector. Loop over the vectors to plot. (Will try to give implementation, as soon as I have my laptop)

user_135594740
Related