np.savez() for multiple arrays in NumPy - Step-by-Step Execution

Concept Flow - np.savez() for multiple arrays

Prepare arrays to save

↓

Call np.savez(filename, arrays...)

↓

Arrays packed into a .npz file

↓

File saved on disk

↓

Later: Load .npz file with np.load()

↓

Access arrays by keys

↓

Use arrays in your program

You prepare arrays, save them together in one .npz file, then load and access them later by name.

Execution Sample

NumPy

import numpy as np

arr1 = np.array([1, 2, 3])
arr2 = np.array([4, 5, 6])

np.savez('data.npz', first=arr1, second=arr2)

This code saves two arrays named 'first' and 'second' into one file called 'data.npz'.

Execution Table

Step	Action	Input Arrays	File Created	Output
1	Create arr1	[1, 2, 3]	No	arr1 ready
2	Create arr2	[4, 5, 6]	No	arr2 ready
3	Call np.savez('data.npz', first=arr1, second=arr2)	[arr1, arr2]	Yes	File 'data.npz' with arrays saved
4	Load file with np.load('data.npz')	File 'data.npz'	No	Loaded file object with keys 'first', 'second'
5	Access arrays by keys	Loaded file	No	Arrays retrieved: first=[1 2 3], second=[4 5 6]

💡 All arrays saved and accessible from the .npz file.

Variable Tracker

Variable	Start	After Step 1	After Step 2	After Step 3	After Step 4	After Step 5
arr1	None	[1 2 3]	[1 2 3]	[1 2 3]	[1 2 3]	[1 2 3]
arr2	None	None	[4 5 6]	[4 5 6]	[4 5 6]	[4 5 6]
file 'data.npz'	Not created	Not created	Created	Created	Loaded	Loaded
loaded_file	None	None	None	None	File object	File object
loaded_file['first']	None	None	None	None	None	[1 2 3]
loaded_file['second']	None	None	None	None	None	[4 5 6]

Key Moments - 3 Insights

Why do we give names like 'first' and 'second' when saving arrays?

Because np.savez saves arrays with keys. These names become keys to access arrays later, as shown in execution_table step 5.

What happens if we don't give names to arrays in np.savez()?

Arrays get default names like 'arr_0', 'arr_1'. You can still access them by these keys after loading.

Is the .npz file human-readable?

No, it's a compressed binary file storing arrays efficiently. You must load it with np.load to read arrays.

Visual Quiz - 3 Questions

Look at the execution_table at step 3, what is created?

A A new array in memory

B A text file with array data

C A file named 'data.npz' containing arrays

D An error because arrays are unnamed

💡 Hint

Check the 'File Created' and 'Output' columns at step 3 in execution_table.

According to variable_tracker, what is the value of loaded_file['second'] after step 5?

A [4 5 6]

B None

C [1 2 3]

D File object

💡 Hint

Look at the last column for loaded_file['second'] in variable_tracker.

If we omit names in np.savez(arr1, arr2), how do we access the second array after loading?

A loaded_file['second']

B loaded_file['arr_1']

C loaded_file[1]

D loaded_file['arr2']

💡 Hint

Recall key naming defaults mentioned in key_moments about unnamed arrays.

Concept Snapshot

np.savez(filename, **arrays) saves multiple arrays in one .npz file.
Each array is saved with a key (name).
Load with np.load(filename) to get a file object.
Access arrays by keys like loaded['keyname'].
Useful to store and share multiple arrays together.

Full Transcript

This visual execution shows how np.savez saves multiple numpy arrays into a single .npz file. First, arrays are created in memory. Then np.savez is called with a filename and named arrays. This creates a compressed file storing all arrays with their keys. Later, np.load reads the file and returns an object to access arrays by their keys. The execution table traces each step from array creation, saving, loading, to accessing arrays. The variable tracker shows how variables and file states change. Key moments clarify why naming arrays is important and that .npz files are binary. The quiz tests understanding of file creation, array access, and default naming. The snapshot summarizes usage in a few lines.