How to remove array rows that contain only 0 using NumPy?

Numpy library provides a function called numpy.all() that returns True when all elements of n-d array passed to the first parameter are True else it returns False. Thus, to determine the entire row containing 0’s can be removed by specifying axis=1. It will traverse each row and will check for the condition given in first parameter. 

Example:

data=[[1,2,3]
      [0,0,0]
      [9,8,7]]
After removing row with all zeroes:
data=[[1,2,3]
      [9,8,7]]

Example 1:

Approach Followed:

• Take a numpy n-d array.
• Remove rows that contain only zeroes using numpy.all() function.
• Print the n-d array.

Python3

import numpy as np
# take data
data = np.array([[1, 2, 3], [0, 0, 0], [4, 5, 6],
                 [0, 0, 0], [7, 8, 9], [0, 0, 0]])
# print original data having rows with all zeroes
print("Original Dataset")
print(data)

# remove rows having all zeroes
data = data[~np.all(data == 0, axis=1)]

# data after removing rows having all zeroes
print("After removing rows")
print(data)

Output:

Example 2:

Approach Followed:

• Take 20 random numbers between 0-10, using numpy.random.choice() method.
• Align them in rows and columns, using reshape() method.
• Explicitly mark some rows as completely 0.
• Remove rows having all zeroes.
• Print dataset.

Python3

import numpy as np
# take random data

# random.choice(x,y) will pick y elements from range (0,(x-1))
data = np.random.choice(10, 20)

# specify the dimensions of data i.e (rows,columns)
data = data.reshape(5, 4)

# print original data having rows with all zeroes
print("Original Dataset")
print(data)

# make some rows entirely zero
data[1, :] = 0  # making 2nd row entirely 0
data[4, :] = 0  # making last row entirely 0

# after making 2nd and 5th row as 0
print("After making some rows as entirely 0")
print(data)
data = data[~np.all(data == 0, axis=1)]

# data after removing rows having all zeroes
print("After removing rows")
print(data)

Output: