MATLAB file I/O (loadmat, savemat) in SciPy - Practice Problems & Coding Challenges

Challenge - 5 Problems

❓ Predict Output

intermediate

What is the output of loading a MATLAB .mat file with nested structs?

Consider a MATLAB .mat file saved with a nested struct variable. When loaded using scipy.io.loadmat, what is the type of the nested struct in Python?

SciPy

import scipy.io
mat = scipy.io.loadmat('nested_struct.mat')
nested = mat['nested_struct']
type_nested = type(nested[0,0]['field1'][0,0])
print(type_nested)

A <class 'scipy.io.matlab.mio5_params.mat_struct'>

B <class 'dict'>

C <class 'numpy.ndarray'>

D <class 'list'>

💡 Hint

MATLAB structs are loaded as special objects, not plain dicts.

When loading MATLAB structs with loadmat, nested structs become mat_struct objects, which behave like objects with attributes.

❓ data_output

intermediate

How many variables are loaded from a .mat file with 3 variables?

You load a MATLAB file containing exactly 3 variables using scipy.io.loadmat. How many keys will the resulting dictionary have, excluding MATLAB metadata keys?

SciPy

import scipy.io
mat = scipy.io.loadmat('three_vars.mat')
keys = [k for k in mat.keys() if not k.startswith('__')]
print(len(keys))

A 1

B 5

C 0

D 3

💡 Hint

MATLAB metadata keys start and end with double underscores.

The loaded dictionary contains all variables as keys plus metadata keys starting with '__'. Filtering those out leaves exactly the number of variables saved.

🔧 Debug

advanced

Why does saving a dictionary with a list value cause an error in savemat?

Given the code below, why does scipy.io.savemat raise an error?

import scipy.io
data = {'arr': [1, 2, 3]}
scipy.io.savemat('file.mat', data)

SciPy

import scipy.io
data = {'arr': [1, 2, 3]}
scipy.io.savemat('file.mat', data)

A savemat requires all values to be strings

B savemat expects numpy arrays, not Python lists

C savemat cannot save dictionaries

D savemat only saves scalar values

💡 Hint

Check the data type of values passed to savemat.

savemat requires values to be numpy arrays or scalars. Python lists are not accepted and cause a TypeError.

🚀 Application

advanced

How to save a Python dictionary with mixed data types to a .mat file?

You want to save a Python dictionary with keys 'name' (string), 'age' (int), and 'scores' (list of floats) to a MATLAB .mat file. Which approach works correctly?

A Convert 'scores' to numpy array, then use savemat

B Save the dictionary directly without changes

C Convert 'age' to string, then save

D Convert all values to lists, then save

💡 Hint

MATLAB expects arrays or scalars, not Python lists or mixed types.

To save mixed data types, convert lists to numpy arrays and keep strings and ints as is. Then savemat can save the dictionary properly.

🧠 Conceptual

expert

What happens to MATLAB cell arrays when loaded with loadmat?

When loading a MATLAB .mat file containing a cell array using scipy.io.loadmat, how is the cell array represented in Python?

A As a dict with keys as indices

B As a Python list of lists

C As a numpy ndarray with dtype=object containing elements

D As a pandas DataFrame

💡 Hint

Cell arrays hold mixed types and are loaded as arrays with object dtype.

MATLAB cell arrays become numpy ndarrays with dtype=object, where each element corresponds to a cell content.