NumPy RuntimeWarning: invalid value encountered in divide

# Table of Contents

# NumPy RuntimeWarning: invalid value encountered in divide

The NumPy "RuntimeWarning: invalid value encountered in divide" issue occurs when you try to divide by zero, INF or divide by a missing value, such as NaN.

To resolve the issue, use the numpy.seterr() method to specify that invalid division operations should be ignored.

Here is an example of when the warning is shown.

main.py

Copied!
import numpy as np

arr1 = np.array([0, 2, 4, 6, 8])

arr2 = np.array([0, 1, 2, 3, 4])

result = np.divide(arr1, arr2)

# /home/borislav/Desktop/bobbyhadz_python/main.py:17: RuntimeWarning: invalid value encountered in divide
  # np.divide(x, y)
print(result)

numpy runtime warning invalid valid encountered in divide

The first element-wise division tries to divide 0 by 0.

You'd get the same warning if only the divisor is set to 0.

main.py

Copied!
import numpy as np

arr1 = np.array([1, 2, 4, 6, 8])

arr2 = np.array([0, 1, 2, 3, 4])

result = np.divide(arr1, arr2)

# /home/borislav/Desktop/bobbyhadz_python/main.py:7: RuntimeWarning: divide by zero encountered in divide
#   result = np.divide(arr1, arr2)
# [inf  2.  2.  2.  2.]
print(result)

only divisor is set to 0

# Use the `numpy.seterr()` method to resolve the issue

One way to resolve the issue is to use the numpy.seterr() method to specify how floating-point errors and division by zero should be handled.

main.py

Copied!
import numpy as np

np.seterr(divide='ignore', invalid='ignore')

arr1 = np.array([1, 2, 4, 6, 8])

arr2 = np.array([0, 1, 2, 3, 4])


result = np.divide(arr1, arr2)
print(result)

use numpy seterr method to solve the error

The code for this article is available on GitHub

Make sure to use the numpy.seterr() method at the top of your module, before calling numpy.divide().

The numpy.seterr() method doesn't only affect how floating-point errors are handled.

Operations on integer scalar types (such as int16) are also handled like floating-point, so they are also affected by the numpy.seterr() method.

We set the divide keyword argument to "ignore" when calling the method.

main.py

Copied!
import numpy as np

np.seterr(divide='ignore', invalid='ignore')

When the divide argument is set to "ignore", division by zero errors are ignored.

When the invalid argument is set to "ignore", invalid floating-point operations are also ignored.

If you only want to suppress warnings for a block of code, use the with statement with numpy.errstate()

main.py

Copied!
import numpy as np


arr1 = np.array([1, 2, 4, 6, 8])

arr2 = np.array([0, 1, 2, 3, 4])

with np.errstate(divide='ignore', invalid='ignore'):
    result = np.divide(arr1, arr2)
    print(result)

use with statement with numpy errstate

The code for this article is available on GitHub

The numpy.errstate context manager is used for handling floating-point errors.

The settings are only applied to the indented block in the with statement.

Once you exit the with statement, you revert to the default settings.

# Make sure you don't have `inf` values in either of the arrays

Another common cause of the warning is having inf values in one or both of the arrays.

Here is an example.

main.py

Copied!
import numpy as np


arr1 = np.array([1, 2, 4, 6, 8])

arr2 = np.array([np.inf, 1, 2, 3, 4])

result = np.divide(arr1, arr2)


print(np.isfinite(arr1).all()) # 👉️ True

print(np.isfinite(arr2).all()) # 👉️ False

The code for this article is available on GitHub

You can use the numpy.isfinite(arr).all() method to check if all values in the arrays are finite.

The numpy.isfinite() method tests for finiteness (not infinite and not Not a Number).

You can use the numpy.seterr() method if you want to ignore the warning.

main.py

Copied!
import numpy as np

np.seterr(divide='ignore', invalid='ignore')

arr1 = np.array([1, 2, 4, 6, 8])

arr2 = np.array([np.inf, 1, 2, 3, 4])

result = np.divide(arr1, arr2)


print(np.isfinite(arr1).all())

print(np.isfinite(arr2).all())

ignore warning with numpy seterr

The code for this article is available on GitHub

If you only want to disable the warning for a block of code, use the numpy.errstate() method.

main.py

Copied!
import numpy as np

arr1 = np.array([1, 2, 4, 6, 8])

arr2 = np.array([np.inf, 1, 2, 3, 4])

with np.errstate(divide='ignore', invalid='ignore'):
    result = np.divide(arr1, arr2)
    print(result)


print(np.isfinite(arr1).all())

print(np.isfinite(arr2).all())

disable warning for block of code

Ignoring the warning is OK most of the time as there are many cases where division by zero is not an issue because it is handled in a later operation.

# Additional Resources

You can learn more about the related topics by checking out the following tutorials:

I wrote a book in which I share everything I know about how to become a better, more efficient programmer.

You can use the search field on my Home Page to filter through all of my articles.

NumPy RuntimeWarning: invalid value encountered in divide

# Table of Contents