Controlling copy behavior in NumPy

Introduction

Sometimes you want to change data without making a full copy to save memory. Other times, you want a separate copy to avoid changing the original data by mistake.

When you want to modify an array but keep the original unchanged.

When you want to save memory by working with views instead of copies.

When you want to pass data to a function without copying it.

When you want to create a backup of data before changing it.

When you want to avoid unexpected changes due to shared data.

Syntax

NumPy

array.copy()
array.view()

copy() makes a full separate copy of the array data.

view() creates a new array object that looks at the same data (no copy).

Examples

This shows how to make a full copy and a view of the same array.

NumPy

import numpy as np

arr = np.array([1, 2, 3])
arr_copy = arr.copy()  # full copy
arr_view = arr.view()  # view, no copy

Changing arr_copy does not affect arr, but changing arr_view does.

NumPy

arr_copy[0] = 100
arr_view[1] = 200

print(arr)  # original array

Sample Program

This program shows how changing the copy does not affect the original array, but changing the view does affect it because they share the same data.

NumPy

import numpy as np

# Original array
arr = np.array([10, 20, 30])

# Make a full copy
arr_copy = arr.copy()

# Make a view (no copy)
arr_view = arr.view()

# Change the copy
arr_copy[0] = 999

# Change the view
arr_view[1] = 888

print('Original array:', arr)
print('Copy array:', arr_copy)
print('View array:', arr_view)

Important Notes

Use copy() when you want to protect the original data from changes.

Use view() to save memory and speed when you only need to look at or temporarily change data.

Be careful: modifying a view changes the original array, which can cause bugs if unexpected.

Summary

copy() makes a full separate copy of the array data.

view() creates a new array object sharing the same data (no copy).

Choose copy or view depending on whether you want to protect original data or save memory.