如何绘制此图？

Question

Hi Data Visulization Experts!

I'm trying to plot this graph hand-drawn example here in python.

However, it seems it will be much trickier to plot as compared to normal plots.

It would be really nice if a visualisation expert can help in plotting this. In the hand-drawn figure I kept all the lines red, however, they can be in different colours.

Thank you for your responses :)

Answer 1

Without the data I cannot assist in plotting it, but you can solve it via Python by following the examples at:

https://matplotlib.org/gallery/units/bar_demo2.html?highlight=bar

^ You can perhaps have each bar chart on a separate subplot?

If your input is multiple sets of data you can look into maybe a multiple dataset histogram at

https://matplotlib.org/gallery/statistics/histogram_multihist.html?highlight=hist

^ If you want to have the first column of all categories next to each other, then the second bar, etc..

Perhaps you would prefer just a standard histogram though:

https://matplotlib.org/gallery/statistics/histogram_features.html?highlight=hist

It really depends on what you are trying to emphasize from your data

@EDIT

Here is a Python program to plot your information! You can easily adjust the width (I commented out a way to set the width to cover the empty gaps, but its slightly off) and to plot a line I would use width = 0.01

import matplotlib.pyplot as plt
import numpy as np

data = {0: [4, 8, 6],
        1: [2, 4, 3, 6],
        2: [3, 6],
        3: [10, 3, 8, 6, 10, 12]}

fig, ax = plt.subplots(tight_layout=True)

for key, values in data.items():
    ind = np.arange(key, key + 1, 1/len(data[key]))
    width = 0.1 # 1/len(data[key])
    ax.bar(ind + width/len(values), data[key], width, align='center', label="Category {}".format(key + 1))

ax.xaxis.set_ticks([])
ax.set_xticklabels(' ')
ax.legend()
ax.set_ylabel('Some vertical label')
plt.title('My plot for Stack Overflow')
plt.show()

This outputs: Which I think is what you're looking for. This may not be a perfect solution, but I think it showcases the methodology used to make such plots :). If this solution solved your problem, I would appreciate if you could click the check mark by my post to accept it as the answer you were looking for!

Answer 2

Set your data up like this, then after you've selected the range for a bar chart, you'll need to swap the X and Y axis: