How to create and plot XOR data for a specific range in Python?

Question

I want to create such data like in the image below

Desired image

To do this I used this code;

centers = [[-2,2],[-2,-2],[2,-2],[2,2]]
X, y = make_blobs(n_samples = 400, n_features=2, centers=centers, cluster_std=0.8, random_state=40)

Then I splited X into two part

X1 = X[:,0]
X2 = X[:,1]

When plot this data I got this image

plt.scatter(X1,y,color="green")
plt.scatter(X2,y,color="red")

What I got

Is there a way to create such data, or is there any idea about what I did wrong?

Answer 1

Your y hold the integer labels of the categories, which are 0,1,2 or 3 in your example. So you are plotting x against those four values. Split the X (which holds the x,y coord) like so:

import matplotlib.pyplot as plt
from sklearn.datasets import make_blobs

centers = [[-2,2],[-2,-2],[2,-2],[2,2]]
X, y = make_blobs(n_samples = 400, n_features=2, centers=centers, cluster_std=0.8, random_state=40)

X1 = [x for idx,x in enumerate(X[:,0]) if y[idx]<=2]
Y1 = [x for idx,x in enumerate(X[:,1]) if y[idx]<=2]
X2 = [x for idx,x in enumerate(X[:,0]) if y[idx]>=3]
Y2 = [x for idx,x in enumerate(X[:,1]) if y[idx]>=3]

plt.scatter(X1,Y1,color="green")
plt.scatter(X2,Y2,color="red")

plt.show()

which gives (you can use different criteria for the split, just an example):

Answer 2

To create and plot the XOR dataset between a specific range, I wrote this code:

import numpy as np
import matplotlib.pyplot as plt

x = np.random.uniform(low=-4, high=4, size=(400,2))
y = np.bitwise_xor(np.sign(x[:,0]).astype(int),np.sign(x[:,1]).astype(int))
plt.scatter(x[:,0],x[:,1],c=y)
plt.xlabel('X1')
plt.ylabel('X2')

plt.show()

As you can see I set an arbitrary interval and I used the .bitwise_xor to calculate the XOR. Notice that I used the .astype(int) attribute to cast the random float numbers to integers.

Which gives this plot