Numpy – Check if Array is all NaN

To check if all the values in a Numpy array are NaN or not, you can use a combination of the numpy.isnan() function and the all() function. The idea is to check if each value in the array is nan or not using numpy.isnan() which results in a boolean array and check if all the values in the resulting boolean array are True or not using the all function.

The following is the syntax –

import numpy as np

# check if numpy array is all zero
np.isnan(ar).all()

Alternatively, you can iterate through the array and return False if you encounter any non-NaN value.

Let’s now look at the methods mentioned above with the help of some examples.

Example 1 – Check if the Array is all NaN using all() function

If you compare an array with a scaler value, the resulting array would be a boolean array with True for the array values that were equal to the scaler value and False otherwise.

In this method, we compare the array with np.nan and check if all the values in the resulting boolean array are True or not using the numpy array all() function.

Let’s look at an example

import numpy as np

# create two arrays
ar1 = np.array([np.nan, np.nan, np.nan, np.nan])
ar2 = np.array([np.nan, 0, 1, np.nan])

# check if array is all NaN
print(np.isnan(ar1).all())
print(np.isnan(ar2).all())

Output:

True
False

Here, we created two arrays. ar1 with all values as nan and ar2 with only some values as nan (not all) and then checked if these arrays are all nan or not using our method.

We get True for ar1 and False for ar2, which is the correct result.

Example 2 – Iterate through the array

Alternatively, we can iterate through the entire array, element by element, and check if each element is nan or not using the numpy.isnan() function. If we encounter a non-nan element, we return False.

import numpy as np

# create two arrays
ar1 = np.array([np.nan, np.nan, np.nan, np.nan])
ar2 = np.array([np.nan, 0, 1, np.nan])

# function to check if array is all nan
def is_array_all_nan(ar):
    if len(ar) == 0:
        return False
    for val in ar:
        if np.isnan(val):
            continue
        else:
            return False
    return True

# check if array is all zero
print(is_array_all_nan(ar1))
print(is_array_all_nan(ar2))

Output:

True
False

We get the same result as above. You can think of this method as a more verbose version (and less optimized) version of method 1.

You might also be interested in –

• Numpy – Check if Array is all Zero
• Numpy – Check If Array has any Duplicates
• How to search and replace text in a file using Python?