np.genfromtxt() for handling missing data in NumPy - Practice Problems & Coding Challenges

Challenge - 5 Problems

❓ Predict Output

intermediate

Output of np.genfromtxt() with missing values

What is the output of the following code snippet that uses np.genfromtxt() to read data with missing values?

NumPy

import numpy as np
from io import StringIO

data = """
1,2,3
4,,6
7,8,
"""

arr = np.genfromtxt(StringIO(data), delimiter=',', filling_values=-1)
print(arr)

A ValueError: could not convert string to float: ''

B

[[ 1.  2.  3.]
 [ 4. -1.  6.]
 [ 7.  8. -1.]]

C

[[1. 2. 3.]
 [4. nan 6.]
 [7. 8. nan]]

D

[[1 2 3]
 [4 0 6]
 [7 8 0]]

💡 Hint

Look at how filling_values replaces missing entries.

The filling_values=-1 argument replaces missing data with -1. So missing entries become -1 instead of NaN or errors.

🧠 Conceptual

intermediate

Understanding the role of filling_values in np.genfromtxt()

What does the filling_values parameter do in np.genfromtxt() when reading a file with missing data?

A It specifies the value to replace missing data entries during loading.

B It skips rows that contain any missing data.

C It converts all data to strings instead of numbers.

D It raises an error if any missing data is found.

💡 Hint

Think about how missing data can be handled automatically.

The filling_values parameter tells np.genfromtxt() what value to use instead of missing entries, so the array remains numeric and complete.

❓ data_output

advanced

Resulting array shape and content with missing data

Given this CSV data with missing values, what is the shape and content of the numpy array after loading with np.genfromtxt() using delimiter=',' and default parameters?

NumPy

import numpy as np
from io import StringIO

data = """
10,20,30
40,,60
,80,90
"""

arr = np.genfromtxt(StringIO(data), delimiter=',')
print(arr)
print(arr.shape)

A

[[10. 20. 30.]
 [40. 0. 60.]
 [0. 80. 90.]]
(3, 3)

B

[[10 20 30]
 [40 0 60]
 [0 80 90]]
(3, 3)

C ValueError: could not convert string to float: ''

D

[[10. 20. 30.]
 [40. nan 60.]
 [nan 80. 90.]]
(3, 3)

💡 Hint

By default, missing values become NaN in float arrays.

Without specifying filling_values, missing entries become nan and the array is float type with shape (3,3).

🔧 Debug

advanced

Identify the error when loading CSV with missing data

What error will this code raise when trying to load CSV data with missing values using np.genfromtxt() without specifying filling_values?

NumPy

import numpy as np
from io import StringIO

data = """
1,2,3
4,,6
7,8,9
"""

arr = np.genfromtxt(StringIO(data), delimiter=',', dtype=int)
print(arr)

A No error, prints array with zeros for missing values

B TypeError: unsupported operand type(s) for +: 'int' and 'str'

C ValueError: invalid literal for int() with base 10: ''

D SyntaxError: invalid syntax

💡 Hint

Missing values cannot be converted to int without filling.

Missing entries are empty strings which cannot convert to int, causing a ValueError.

🚀 Application

expert

Handling mixed missing data types with np.genfromtxt()

You have a CSV file with numeric and string columns, some missing values in both. Which np.genfromtxt() call correctly loads the data, replacing missing numeric values with -999 and missing strings with 'missing'?

NumPy

import numpy as np
from io import StringIO

data = """
1,apple,3.5
2,,
,banana,4.1
"""

A np.genfromtxt(StringIO(data), delimiter=',', dtype=None, encoding=None, filling_values={0: -999, 1: 'missing', 2: -999})

B np.genfromtxt(StringIO(data), delimiter=',', dtype=None, encoding=None, filling_values=-999)

C np.genfromtxt(StringIO(data), delimiter=',', dtype='U10,f8,i4', filling_values=['missing', -999, -999])

D np.genfromtxt(StringIO(data), delimiter=',', dtype=None, encoding=None, missing_values='', filling_values='missing')

💡 Hint

Use a dictionary for filling_values to specify per-column replacements.

Using a dictionary for filling_values allows specifying different fill values per column, matching their data types.