Sparse matrix file I/O in SciPy - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to save a sparse matrix to a file using scipy.

SciPy

from scipy import sparse

matrix = sparse.csr_matrix([[0, 0, 1], [1, 0, 0]])
sparse.[1]('matrix.npz', matrix)

Drag options to blanks, or click blank then click option'

Aload_npz

Bload

Csave

Dsave_npz

Attempts:

3 left

💡 Hint

Common Mistakes

Using load_npz instead of save_npz.

Using generic save instead of save_npz.

The save_npz function saves a sparse matrix to a file in .npz format.

2fill in blank

medium

Complete the code to load a sparse matrix from a file using scipy.

SciPy

from scipy import sparse

loaded_matrix = sparse.[1]('matrix.npz')

Drag options to blanks, or click blank then click option'

Aload_npz

Bsave_npz

Cload

Dsave

Attempts:

3 left

💡 Hint

Common Mistakes

Using save_npz instead of load_npz.

Using generic load instead of load_npz.

The load_npz function loads a sparse matrix from a .npz file.

3fill in blank

hard

Fix the error in the code to correctly save a sparse matrix to a file.

SciPy

from scipy import sparse

matrix = sparse.csr_matrix([[1, 0], [0, 1]])
sparse.save_npz('data.npz', [1])

Drag options to blanks, or click blank then click option'

Amatrix.todense()

Bmatrix.toarray()

Cmatrix

Dmatrix.data

Attempts:

3 left

💡 Hint

Common Mistakes

Passing a dense array instead of a sparse matrix.

Passing the data attribute of the sparse matrix.

The save_npz function requires a sparse matrix, not a dense array or data attribute.

4fill in blank

hard

Fill both blanks to save and then load a sparse matrix correctly.

SciPy

from scipy import sparse

matrix = sparse.csr_matrix([[0, 2], [3, 0]])
sparse.[1]('file.npz', matrix)
loaded = sparse.[2]('file.npz')

Drag options to blanks, or click blank then click option'

Asave_npz

Bload_npz

Csave

Dload

Attempts:

3 left

💡 Hint

Common Mistakes

Mixing save and load functions incorrectly.

Using generic save or load instead of npz-specific functions.

Use save_npz to save and load_npz to load sparse matrices.

5fill in blank

hard

Fill all three blanks to create, save, and load a sparse matrix, then convert it to dense.

SciPy

from scipy import sparse

matrix = sparse.[1]([[1, 0], [0, 1]])
sparse.[2]('matrix.npz', matrix)
loaded = sparse.[3]('matrix.npz')
dense = loaded.toarray()

Drag options to blanks, or click blank then click option'

Acsr_matrix

Bsave_npz

Cload_npz

Dcsc_matrix

Attempts:

3 left

💡 Hint

Common Mistakes

Using csc_matrix instead of csr_matrix for creation.

Using generic save/load instead of save_npz/load_npz.

Create the sparse matrix with csr_matrix, save with save_npz, and load with load_npz.

Practice

1. What is the main purpose of using save_npz and load_npz functions in scipy.sparse?

easy

A. To perform matrix multiplication on sparse matrices

B. To convert sparse matrices into dense matrices

C. To visualize sparse matrices as heatmaps

D. To save and load sparse matrices efficiently without losing their structure

Solution

1. Step 1: Understand the purpose of save_npz and load_npz

   These functions are designed to save sparse matrices to disk and load them back while keeping their sparse format intact.

2. Step 2: Compare options with the purpose

   Only To save and load sparse matrices efficiently without losing their structure correctly describes saving and loading sparse matrices efficiently without losing their sparse structure.

3. Final Answer:

   To save and load sparse matrices efficiently without losing their structure -> Option D

4. Quick Check:

   Sparse matrix file I/O = save/load sparse matrices [OK]

Hint: Remember save_npz/load_npz keep sparse format intact [OK]

Common Mistakes:

• Thinking these functions convert to dense matrices
• Confusing file I/O with matrix operations
• Assuming visualization is part of file I/O

2. Which of the following is the correct way to save a sparse matrix sp_matrix to a file named data.npz using SciPy?

easy

A. scipy.sparse.save('data.npz', sp_matrix)

B. scipy.sparse.save_npz('data.npz', sp_matrix)

C. scipy.sparse.load_npz('data.npz', sp_matrix)

D. scipy.save_npz(sp_matrix, 'data.npz')

Solution

1. Step 1: Identify the correct function and argument order

   The function to save sparse matrices is save_npz from scipy.sparse, and it takes the filename first, then the matrix.

2. Step 2: Check each option

   scipy.sparse.save_npz('data.npz', sp_matrix) matches the correct syntax: save_npz('filename', matrix). Others either use wrong function names or argument order.

3. Final Answer:

   scipy.sparse.save_npz('data.npz', sp_matrix) -> Option B

4. Quick Check:

   save_npz(filename, matrix) = scipy.sparse.save_npz('data.npz', sp_matrix) [OK]

Hint: save_npz(filename, matrix) saves sparse matrix [OK]

Common Mistakes:

• Using load_npz instead of save_npz to save
• Swapping filename and matrix arguments
• Using non-existent save function

3. Consider the following code snippet:

from scipy.sparse import csr_matrix, save_npz, load_npz
import numpy as np

arr = np.array([[0, 0, 1], [1, 0, 0], [0, 2, 0]])
sp = csr_matrix(arr)
save_npz('matrix.npz', sp)
loaded_sp = load_npz('matrix.npz')
print(loaded_sp.toarray())

What will be the output printed?

medium

A. Error: cannot convert sparse matrix to array

B. [[0 0 0]
[0 0 0]
[0 0 0]]

C. [[0 0 1]
[1 0 0]
[0 2 0]]

D. [[1 0 0]
[0 1 0]
[0 0 1]]

Solution

1. Step 1: Create sparse matrix and save it

   The code creates a sparse matrix from the numpy array, saves it to 'matrix.npz', then loads it back.

2. Step 2: Convert loaded sparse matrix to dense array and print

   Using toarray() converts the sparse matrix back to the original dense numpy array, so the printed output matches the original array.

3. Final Answer:

   [[0 0 1] [1 0 0] [0 2 0]] -> Option C

4. Quick Check:

   load_npz + toarray() = original array [OK]

Hint: load_npz returns sparse; use toarray() to see full matrix [OK]

Common Mistakes:

• Expecting zeros after loading
• Forgetting to convert sparse to dense before printing
• Confusing save_npz and load_npz usage

4. You wrote this code to load a sparse matrix:

from scipy.sparse import load_npz
matrix = load_npz('data.npz')
print(matrix)

But you get an error: ModuleNotFoundError: No module named 'scipy.sparse'. What is the most likely cause?

medium

A. You forgot to install the SciPy library in your environment

B. The file 'data.npz' does not exist

C. You used load_npz instead of save_npz

D. You need to convert the matrix to dense before printing

Solution

1. Step 1: Analyze the error message

   The error says the module 'scipy.sparse' is not found, which means SciPy is not installed or not accessible.

2. Step 2: Check other options

   File missing causes a different error, wrong function usage or printing sparse matrix won't cause module import error.

3. Final Answer:

   You forgot to install the SciPy library in your environment -> Option A

4. Quick Check:

   ModuleNotFoundError = missing SciPy install [OK]

Hint: ModuleNotFoundError means missing package install [OK]

Common Mistakes:

• Assuming file missing causes import error
• Confusing function usage errors with import errors
• Thinking sparse matrix print needs conversion to avoid import error

5. You have a large sparse matrix stored in large_matrix.npz. You want to load it, add 5 to all non-zero elements, and save it back without converting to a dense matrix (to save memory). Which code snippet correctly does this?

hard

A. from scipy.sparse import load_npz, save_npz
mat = load_npz('large_matrix.npz')
mat.data += 5
save_npz('large_matrix.npz', mat)

B. from scipy.sparse import load_npz, save_npz
mat = load_npz('large_matrix.npz')
mat = mat.toarray() + 5
save_npz('large_matrix.npz', mat)

C. import numpy as np
mat = np.load('large_matrix.npz')
mat += 5
np.save('large_matrix.npz', mat)

D. from scipy.sparse import load_npz, save_npz
mat = load_npz('large_matrix.npz')
mat.toarray() += 5
save_npz('large_matrix.npz', mat)

Solution

1. Step 1: Load sparse matrix and modify non-zero elements

   Using mat.data accesses the non-zero values directly. Adding 5 to mat.data updates only those values without converting to dense.

2. Step 2: Save the updated sparse matrix back

   Using save_npz saves the modified sparse matrix efficiently.

3. Step 3: Check other options for correctness

   from scipy.sparse import load_npz, save_npz mat = load_npz('large_matrix.npz') mat = mat.toarray() + 5 save_npz('large_matrix.npz', mat) converts to dense explicitly, adds 5 to all elements including zeros, wastes memory, and save_npz fails on dense array. import numpy as np mat = np.load('large_matrix.npz') mat += 5 np.save('large_matrix.npz', mat) uses numpy load/save which does not handle sparse matrices. from scipy.sparse import load_npz, save_npz mat = load_npz('large_matrix.npz') mat.toarray() += 5 save_npz('large_matrix.npz', mat) tries to add 5 to a dense array but does not assign back, and wastes memory.

4. Final Answer:

   from scipy.sparse import load_npz, save_npz mat = load_npz('large_matrix.npz') mat.data += 5 save_npz('large_matrix.npz', mat) -> Option A

5. Quick Check:

   Modify mat.data for sparse update [OK]

Hint: Change mat.data to update non-zero sparse values [OK]

Common Mistakes:

• Adding scalar directly to sparse matrix (converts to dense)
• Using numpy load/save for sparse matrices
• Modifying dense array without saving changes