How to plot one to many plot using matplotlib

Question

Update: Removing the screenshot, Below is the code from the screenshot:

import numpy as np
import matplotlib.pyplot as plt

x = np.arange(5)
y = np.array([[1,2],[1,2,3],[1,2,3,4],[1,2,3,4,5],[1,2,5,7,9]])

plt.plot(x,y) #gives ValueError: setting an array element with a sequence.

#A relaistic example
age = [20,30,40,50,60]
salary = np.array([[200,350,414],[300,500,612,700],[500,819],[900,1012],[812,712]])

plt.plot(age,salary) #gives ValueError: setting an array element with a sequence.

I am having two arrays each of size 5, elements of y are arrays, and I want them to be plotted against each x, for example at x = 0, I want to plot all the points from y[0], is there a way?

Update: Added another example above to show a realistic case , where I need to plot different salaries of different age people, each age people can have more than one salary.

What do you want this to look like? "I want to plot" is pretty vague. — Jody Klymak, Nov 30 '20 at 15:52
At X=0 I want two dots one at y=1 and another at y=2, similarly at X=1, I want three dots at y=1,y=2,y=3 and so on. In some places I saw such graphs , for example plotting all the salaries of different age people and then there will be a line showing average value at each points. — Devashish Priyadarshi, Dec 01 '20 at 02:24

Mr. T · Accepted Answer · 2020-12-01T03:27:23.290

1

List comprehension to the rescue!

import numpy as np
import matplotlib.pyplot as plt

age = [20,30,40,50,60]
salary = np.array([[200,350,414],[300,500,612,700],[500,819],[900,1012],[812,712]])

#creating x-y tuples
xy = [(k, j) for i, k in enumerate(age) for j in salary[i]]

#unpacking the tuples with zip
plt.scatter(*zip(*xy))

plt.show()

Sample output:

However, irregular numpy arrays should not be created, and this example works perfectly well with a normal list. Just saying.

edited Dec 01 '20 at 03:27

answered Dec 01 '20 at 03:13

Mr. T

11,960
10
32
54

I populated age and salary with random data, and your code performed more than 5 times better than mine where I was calling plot() method as many times as many points were there. age = np.arange(200) salary = [] for a in age: sz = np.random.randint(20) arr = [] for s0 in np.arange(sz): data = np.random.randint(20000) arr.append(data) salary.append(arr) – Devashish Priyadarshi Dec 01 '20 at 12:16
You can look at list comprehensions as an optimized version of a for-loop. I linked to some explanation on list comprehension; you should familiarize yourself with this great Python tool. – Mr. T Dec 01 '20 at 12:28

score 0 · Answer 2 · answered Dec 01 '20 at 03:11

0

As of now I am using the following workaround, but looking for a simpler solution:

indx = -1
for a in age:
    indx+=1
    for s in salary[indx]:
        plt.plot(a,s,'o')
plt.show()

answered Dec 01 '20 at 03:11

Devashish Priyadarshi

105
7

What's wrong with this approach? I mean, you should specify the color but apart from that? – Mr. T Dec 01 '20 at 03:32
In this approach , I am calling the plot method as many times as many points are there to be plotted , hence taking more time , I tested with a dataset of 200 , in x , axis, and mine was taking 5 sec + while yours took few milliseconds, Thanks – Devashish Priyadarshi Dec 01 '20 at 12:13

How to plot one to many plot using matplotlib

2 Answers2