NumPy ndarray.copy() Method

NumPy ndarray.copy()

The numpy.ndarray.copy() method returns a copy of the array. This function is useful when you need to create an independent copy of an array rather than a reference to the original data.

Syntax

ndarray.copy(order='C')

Parameters

Parameter	Type	Description
order	str, optional	Memory layout of the copy. Options: 'C': Row-major (C-style) order. 'F': Column-major (Fortran-style) order. 'A': Preserve original array’s order. 'K': Preserve array’s layout as much as possible. Default is 'C'.

Return Value

Returns a new ndarray object that is a copy of the original array. The copy is independent of the original array and does not share memory with it.

---

Examples

1. Creating a Copy of an Array

In this example, we create an array and make a copy using ndarray.copy(). We then modify the original array to show that the copied array remains unchanged.

import numpy as np

# Create an original array
original_array = np.array([1, 2, 3, 4, 5])

# Create a copy of the array
copied_array = original_array.copy()

# Modify the original array
original_array[0] = 99

# Print both arrays
print("Original Array:", original_array)
print("Copied Array:", copied_array)  # Should remain unchanged

Output:

Original Array: [99  2  3  4  5]
Copied Array:   [1  2  3  4  5]

Since ndarray.copy() creates an independent copy, changes to original_array do not affect copied_array.

2. Copying an Array in Fortran Order

We can use the order='F' parameter to create a copy with Fortran-style (column-major) memory layout.

import numpy as np

# Create a 2D array
original_array = np.array([[1, 2], [3, 4]])

# Create a Fortran-order copy
copied_array = original_array.copy(order='F')

# Print memory layout of both arrays
print("Original Array Flags:\n", original_array.flags)
print("Copied Array Flags (Fortran Order):\n", copied_array.flags)

Output:

Original Array Flags:
   C_CONTIGUOUS : True
  F_CONTIGUOUS : False
  OWNDATA : True
  WRITEABLE : True
  ALIGNED : True
  WRITEBACKIFCOPY : False

Copied Array Flags (Fortran Order):
   C_CONTIGUOUS : False
  F_CONTIGUOUS : True
  OWNDATA : True
  WRITEABLE : True
  ALIGNED : True
  WRITEBACKIFCOPY : False

Here, the copied array is stored in Fortran order, which means it is optimized for column-major access patterns.

3. Checking If Copy Shares Memory with Original

Using np.may_share_memory(), we can check whether two arrays share the same memory location.

import numpy as np

# Create an array
original_array = np.array([10, 20, 30, 40])

# Make a copy
copied_array = original_array.copy()

# Check memory sharing
print("Do the arrays share memory?", np.may_share_memory(original_array, copied_array))

Output:

Do the arrays share memory? False

The copied array is completely independent, ensuring that modifications to one do not affect the other.