NumPy ndarray.flat Attribute

NumPy ndarray.flat

The numpy.ndarray.flat attribute returns a 1-D iterator over the elements of a NumPy array, allowing element-wise iteration in a flattened manner.

Syntax

ndarray.flat

Return Value

Returns a numpy.flatiter object, which is an iterator that enables iteration over a multi-dimensional array as if it were 1-D.

---

Examples

1. Iterating Over Elements Using ndarray.flat

We create a 2D array and use flat to iterate through its elements as if it were a 1D array.

import numpy as np

# Creating a 2D NumPy array
arr = np.array([[1, 2, 3], 
                [4, 5, 6]])

# Using the flat attribute to iterate through all elements
for element in arr.flat:
    print(element)

Output:

1
2
3
4
5
6

Although the original array is 2D, flat allows us to iterate over it as if it were a 1D array.

2. Accessing Elements Using Indices with ndarray.flat

The flat iterator allows direct indexing to access elements in a flattened array.

import numpy as np

# Creating a 3x3 NumPy array
arr = np.array([[10, 20, 30], 
                [40, 50, 60], 
                [70, 80, 90]])

# Accessing elements using flat iterator
print(arr.flat[0])  # First element (10)
print(arr.flat[4])  # Fifth element (50)
print(arr.flat[-1]) # Last element (90)

Output:

10
50
90

Using indices, we can access elements in a flattened view of the array, even if it has multiple dimensions.

3. Modifying Elements Using ndarray.flat

We can also use flat to modify elements directly in the original array.

import numpy as np

# Creating a 2D NumPy array
arr = np.array([[1, 2, 3], 
                [4, 5, 6]])

# Modifying elements using flat
arr.flat[1] = 99   # Changing second element
arr.flat[-1] = 88  # Changing last element

print(arr)

Output:

[[ 1 99  3]
 [ 4  5 88]]

Even though we modify elements using flat (as if the array were 1D), the changes reflect in the original multi-dimensional structure.

4. Using ndarray.flat in Slicing

We can assign values to multiple elements at once using slicing in flat.

import numpy as np

# Creating a 2D NumPy array
arr = np.array([[1, 2, 3], 
                [4, 5, 6]])

# Assigning values using flat slicing
arr.flat[1:4] = [99, 77, 55]  # Modify elements at indices 1, 2, and 3

print(arr)

Output:

[[ 1 99 77]
 [55  5  6]]

Using slicing, we update multiple elements efficiently in a multi-dimensional array through a flattened view.