🕷️ Crawler Inspector

[](https://www.delftstack.com/) - [Tutorials](https://www.delftstack.com/tutorial/) - [HowTos](https://www.delftstack.com/howto/) - [Reference](https://www.delftstack.com/api/) - [en](https://www.delftstack.com/howto/numpy/numpy-remove-nan-values/) - [DE](https://www.delftstack.com/de/howto/numpy/numpy-remove-nan-values/) - [ES](https://www.delftstack.com/es/howto/numpy/numpy-remove-nan-values/) - [FR](https://www.delftstack.com/fr/howto/numpy/numpy-remove-nan-values/) - [IT](https://www.delftstack.com/it/howto/numpy/numpy-remove-nan-values/) - [JA](https://www.delftstack.com/ja/howto/numpy/numpy-remove-nan-values/) - [KO](https://www.delftstack.com/ko/howto/numpy/numpy-remove-nan-values/) - [PT](https://www.delftstack.com/pt/howto/numpy/numpy-remove-nan-values/) - [ZH](https://www.delftstack.com/zh/howto/numpy/numpy-remove-nan-values/) - [ZH-TW](https://www.delftstack.com/zh-tw/howto/numpy/numpy-remove-nan-values/) [Tutorial](https://www.delftstack.com/tutorial/) [Python 3 Basic](https://www.delftstack.com/tutorial/python-3-basic-tutorial/) [Tkinter](https://www.delftstack.com/tutorial/tkinter-tutorial/) [Python Modules](https://www.delftstack.com/tutorial/python-modules-tutorial/) [JavaScript](https://www.delftstack.com/tutorial/javascript/)[Python Numpy](https://www.delftstack.com/tutorial/python-numpy/)[Git](https://www.delftstack.com/tutorial/git/)[Matplotlib](https://www.delftstack.com/tutorial/matplotlib/)[PyQt5](https://www.delftstack.com/tutorial/pyqt5/)[Data Structure](https://www.delftstack.com/tutorial/data-structure/)[Algorithm](https://www.delftstack.com/tutorial/algorithm/) [HowTo](https://www.delftstack.com/howto/) [Python Scipy](https://www.delftstack.com/howto/python-scipy/)[Python](https://www.delftstack.com/howto/python/)[Python Tkinter](https://www.delftstack.com/howto/python-tkinter/)[Batch](https://www.delftstack.com/howto/batch/)[PowerShell](https://www.delftstack.com/howto/powershell/)[Python Pandas](https://www.delftstack.com/howto/python-pandas/)[Numpy](https://www.delftstack.com/howto/numpy/)[Python Flask](https://www.delftstack.com/howto/python-flask/)[Django](https://www.delftstack.com/howto/django/)[Matplotlib](https://www.delftstack.com/howto/matplotlib/)[Docker](https://www.delftstack.com/howto/docker/)[Plotly](https://www.delftstack.com/howto/plotly/)[Seaborn](https://www.delftstack.com/howto/seaborn/)[Matlab](https://www.delftstack.com/howto/matlab/)[Linux](https://www.delftstack.com/howto/linux/)[Git](https://www.delftstack.com/howto/git/)[C](https://www.delftstack.com/howto/c/)[Cpp](https://www.delftstack.com/howto/cpp/)[HTML](https://www.delftstack.com/howto/html/)[JavaScript](https://www.delftstack.com/howto/javascript/)[jQuery](https://www.delftstack.com/howto/jquery/)[Python Pygame](https://www.delftstack.com/howto/python-pygame/)[TensorFlow](https://www.delftstack.com/howto/tensorflow/)[TypeScript](https://www.delftstack.com/howto/typescript/)[Angular](https://www.delftstack.com/howto/angular/)[React](https://www.delftstack.com/howto/react/)[CSS](https://www.delftstack.com/howto/css/)[PHP](https://www.delftstack.com/howto/php/)[Java](https://www.delftstack.com/howto/java/)[Go](https://www.delftstack.com/howto/go/)[Kotlin](https://www.delftstack.com/howto/kotlin/)[Node.js](https://www.delftstack.com/howto/node.js/)[Csharp](https://www.delftstack.com/howto/csharp/)[Rust](https://www.delftstack.com/howto/rust/)[Ruby](https://www.delftstack.com/howto/ruby/)[Arduino](https://www.delftstack.com/howto/arduino/)[MySQL](https://www.delftstack.com/howto/mysql/)[MongoDB](https://www.delftstack.com/howto/mongodb/)[Postgres](https://www.delftstack.com/howto/postgres/)[SQLite](https://www.delftstack.com/howto/sqlite/)[R](https://www.delftstack.com/howto/r/)[VBA](https://www.delftstack.com/howto/vba/)[Scala](https://www.delftstack.com/howto/scala/)[Raspberry Pi](https://www.delftstack.com/howto/raspberry-pi/) [Reference](https://www.delftstack.com/api/) [Python](https://www.delftstack.com/api/python/)[Python Pandas](https://www.delftstack.com/api/python-pandas/)[Numpy](https://www.delftstack.com/api/numpy/)[Scipy](https://www.delftstack.com/api/scipy/)[JavaScript](https://www.delftstack.com/api/javascript/) # How to Remove Nan Values From a NumPy Array [Vaibhav Vaibhav](https://www.delftstack.com/howto/numpy/numpy-remove-nan-values/#author) Feb 02, 2024 [NumPy](https://www.delftstack.com/tags/numpy/) [NumPy Nan](https://www.delftstack.com/tags/numpy-nan/) 1. [Remove Nan Values Using `logical_not()` and `isnan()` Methods in NumPy](https://www.delftstack.com/howto/numpy/numpy-remove-nan-values/#remove-nan-values-using-logical_not-and-isnan-methods-in-numpy) 2. [Remove Nan Values Using the `isfinite()` Method in NumPy](https://www.delftstack.com/howto/numpy/numpy-remove-nan-values/#remove-nan-values-using-the-isfinite-method-in-numpy) 3. [Remove Nan Values Using the `math.isnan` Method](https://www.delftstack.com/howto/numpy/numpy-remove-nan-values/#remove-nan-values-using-the-mathisnan-method) 4. [Remove Nan Values Using the `pandas.isnull` Method](https://www.delftstack.com/howto/numpy/numpy-remove-nan-values/#remove-nan-values-using-the-pandasisnull-method)  This article will discuss some in-built NumPy functions that you can use to delete `nan` values. ## Remove Nan Values Using `logical_not()` and `isnan()` Methods in NumPy `logical_not()` is used to apply logical `NOT` to elements of an array. `isnan()` is a boolean function that checks whether an element is `nan` or not. Using the `isnan()` function, we can create a boolean array that has `False` for all the non `nan` values and `True` for all the `nan` values. Next, using the `logical_not()` function, We can convert `True` to `False` and vice versa. Lastly, using boolean indexing, We can filter all the non `nan` values from the original NumPy array. All the indexes with `True` as their value will be used to filter the NumPy array. To learn more about these functions in-depth, refer to their [official documentation](https://numpy.org/doc/stable/reference/generated/numpy.logical_not.html#numpy.logical_not) and [here](https://numpy.org/doc/stable/reference/generated/numpy.isnan.html#numpy.isnan), respectively. Refer to the following code snippet for the solution. ``` import numpy as np myArray = np.array([1, 2, 3, np.nan, np.nan, 4, 5, 6, np.nan, 7, 8, 9, np.nan]) output1 = myArray[np.logical_not(np.isnan(myArray))] # Line 1 output2 = myArray[~np.isnan(myArray)] # Line 2 print(myArray) print(output1) print(output2) ``` Output: ``` [ 1. 2. 3. nan nan 4. 5. 6. nan 7. 8. 9. nan] [1. 2. 3. 4. 5. 6. 7. 8. 9.] [1. 2. 3. 4. 5. 6. 7. 8. 9.] ``` `Line 2` is a simplified version of `Line 1`. ## Remove Nan Values Using the `isfinite()` Method in NumPy As the name suggests, the `isfinite()` function is a boolean function that checks whether an element is finite or not. It can also check for finite values in an array and returns a boolean array for the same. The boolean array will store `False` for all the `nan` values and `True` for all the finite values. We will use this function to retrieve a boolean array for the target array. Using boolean indexing, We will filter all the finite values. Again, as mentioned above, indexes with `True` values will be used to filter the array. Here’s the example code. ``` import numpy as np myArray1 = np.array([1, 2, 3, np.nan, np.nan, 4, 5, 6, np.nan, 7, 8, 9, np.nan]) myArray2 = np.array([np.nan, np.nan, np.nan, np.nan, np.nan, np.nan]) myArray3 = np.array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10]) output1 = myArray1[np.isfinite(myArray1)] output2 = myArray2[np.isfinite(myArray2)] output3 = myArray3[np.isfinite(myArray3)] print(myArray1) print(myArray2) print(myArray3) print(output1) print(output2) print(output3) ``` Output: ``` [ 1. 2. 3. nan nan 4. 5. 6. nan 7. 8. 9. nan] [nan nan nan nan nan nan] [ 1 2 3 4 5 6 7 8 9 10] [1. 2. 3. 4. 5. 6. 7. 8. 9.] [] [ 1 2 3 4 5 6 7 8 9 10] ``` To learn more about this function, refer to the [official documentation](https://numpy.org/doc/stable/reference/generated/numpy.isfinite.html#numpy.isfinite) ## Remove Nan Values Using the `math.isnan` Method Apart from these two NumPy solutions, there are two more ways to remove `nan` values. These two ways involve `isnan()` function from `math` library and `isnull` function from `pandas` library. Both these functions check whether an element is `nan` or not and return a boolean result. Here is the solution using `isnan()` method. ``` import numpy as np import math myArray1 = np.array([1, 2, 3, np.nan, np.nan, 4, 5, 6, np.nan, 7, 8, 9, np.nan]) myArray2 = np.array([np.nan, np.nan, np.nan, np.nan, np.nan, np.nan]) myArray3 = np.array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10]) booleanArray1 = [not math.isnan(number) for number in myArray1] booleanArray2 = [not math.isnan(number) for number in myArray2] booleanArray3 = [not math.isnan(number) for number in myArray3] print(myArray1) print(myArray2) print(myArray3) print(myArray1[booleanArray1]) print(myArray2[booleanArray2]) print(myArray3[booleanArray3]) ``` Output: ``` [ 1. 2. 3. nan nan 4. 5. 6. nan 7. 8. 9. nan] [nan nan nan nan nan nan] [ 1 2 3 4 5 6 7 8 9 10] [1. 2. 3. 4. 5. 6. 7. 8. 9.] [] [ 1 2 3 4 5 6 7 8 9 10] ``` ## Remove Nan Values Using the `pandas.isnull` Method Below is the solution using the [`isnull()` method](https://www.delftstack.com/api/python-pandas/pandas-dataframe-dataframe.isnull-and-notnull-function/) from `pandas`. ``` import numpy as np import pandas as pd myArray1 = np.array([1, 2, 3, np.nan, np.nan, 4, 5, 6, np.nan, 7, 8, 9, np.nan]) myArray2 = np.array([np.nan, np.nan, np.nan, np.nan, np.nan, np.nan]) myArray3 = np.array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10]) booleanArray1 = [not pd.isnull(number) for number in myArray1] booleanArray2 = [not pd.isnull(number) for number in myArray2] booleanArray3 = [not pd.isnull(number) for number in myArray3] print(myArray1) print(myArray2) print(myArray3) print(myArray1[booleanArray1]) print(myArray2[booleanArray2]) print(myArray3[booleanArray3]) print(myArray1[~pd.isnull(myArray1)]) # Line 1 print(myArray2[~pd.isnull(myArray2)]) # Line 2 print(myArray3[~pd.isnull(myArray3)]) # Line 3 ``` Output: ``` [ 1. 2. 3. nan nan 4. 5. 6. nan 7. 8. 9. nan] [nan nan nan nan nan nan] [ 1 2 3 4 5 6 7 8 9 10] [1. 2. 3. 4. 5. 6. 7. 8. 9.] [] [ 1 2 3 4 5 6 7 8 9 10] [1. 2. 3. 4. 5. 6. 7. 8. 9.] [] [ 1 2 3 4 5 6 7 8 9 10] ``` Enjoying our tutorials? Subscribe to DelftStack on YouTube to support us in creating more high-quality video guides. [Subscribe](https://www.youtube.com/@delftstack/?sub_confirmation=1) Author: [**Vaibhav Vaibhav**](https://www.delftstack.com/author/vaibhav/) [](https://www.delftstack.com/author/vaibhav/) [](https://www.delftstack.com/author/vaibhav/) Vaibhav is an artificial intelligence and cloud computing stan. He likes to build end-to-end full-stack web and mobile applications. Besides computer science and technology, he loves playing cricket and badminton, going on bike rides, and doodling. Copyright © 2025. All right reserved - [About US](https://www.delftstack.com/about-us/) - [Write For Us](https://www.delftstack.com/write-for-us/) - [Privacy Policy](https://www.delftstack.com/privacy-policy/) - [Advertising](https://www.delftstack.com/advertising/) - [Contact](https://www.delftstack.com/contact/)

1. [Remove Nan Values Using `logical_not()` and `isnan()` Methods in NumPy](https://www.delftstack.com/howto/numpy/numpy-remove-nan-values/#remove-nan-values-using-logical_not-and-isnan-methods-in-numpy) 2. [Remove Nan Values Using the `isfinite()` Method in NumPy](https://www.delftstack.com/howto/numpy/numpy-remove-nan-values/#remove-nan-values-using-the-isfinite-method-in-numpy) 3. [Remove Nan Values Using the `math.isnan` Method](https://www.delftstack.com/howto/numpy/numpy-remove-nan-values/#remove-nan-values-using-the-mathisnan-method) 4. [Remove Nan Values Using the `pandas.isnull` Method](https://www.delftstack.com/howto/numpy/numpy-remove-nan-values/#remove-nan-values-using-the-pandasisnull-method)  This article will discuss some in-built NumPy functions that you can use to delete `nan` values. ## Remove Nan Values Using `logical_not()` and `isnan()` Methods in NumPy `logical_not()` is used to apply logical `NOT` to elements of an array. `isnan()` is a boolean function that checks whether an element is `nan` or not. Using the `isnan()` function, we can create a boolean array that has `False` for all the non `nan` values and `True` for all the `nan` values. Next, using the `logical_not()` function, We can convert `True` to `False` and vice versa. Lastly, using boolean indexing, We can filter all the non `nan` values from the original NumPy array. All the indexes with `True` as their value will be used to filter the NumPy array. To learn more about these functions in-depth, refer to their [official documentation](https://numpy.org/doc/stable/reference/generated/numpy.logical_not.html#numpy.logical_not) and [here](https://numpy.org/doc/stable/reference/generated/numpy.isnan.html#numpy.isnan), respectively. Refer to the following code snippet for the solution. ``` import numpy as np myArray = np.array([1, 2, 3, np.nan, np.nan, 4, 5, 6, np.nan, 7, 8, 9, np.nan]) output1 = myArray[np.logical_not(np.isnan(myArray))] # Line 1 output2 = myArray[~np.isnan(myArray)] # Line 2 print(myArray) print(output1) print(output2) ``` Output: ``` [ 1. 2. 3. nan nan 4. 5. 6. nan 7. 8. 9. nan] [1. 2. 3. 4. 5. 6. 7. 8. 9.] [1. 2. 3. 4. 5. 6. 7. 8. 9.] ``` `Line 2` is a simplified version of `Line 1`. ## Remove Nan Values Using the `isfinite()` Method in NumPy As the name suggests, the `isfinite()` function is a boolean function that checks whether an element is finite or not. It can also check for finite values in an array and returns a boolean array for the same. The boolean array will store `False` for all the `nan` values and `True` for all the finite values. We will use this function to retrieve a boolean array for the target array. Using boolean indexing, We will filter all the finite values. Again, as mentioned above, indexes with `True` values will be used to filter the array. Here’s the example code. ``` import numpy as np myArray1 = np.array([1, 2, 3, np.nan, np.nan, 4, 5, 6, np.nan, 7, 8, 9, np.nan]) myArray2 = np.array([np.nan, np.nan, np.nan, np.nan, np.nan, np.nan]) myArray3 = np.array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10]) output1 = myArray1[np.isfinite(myArray1)] output2 = myArray2[np.isfinite(myArray2)] output3 = myArray3[np.isfinite(myArray3)] print(myArray1) print(myArray2) print(myArray3) print(output1) print(output2) print(output3) ``` Output: ``` [ 1. 2. 3. nan nan 4. 5. 6. nan 7. 8. 9. nan] [nan nan nan nan nan nan] [ 1 2 3 4 5 6 7 8 9 10] [1. 2. 3. 4. 5. 6. 7. 8. 9.] [] [ 1 2 3 4 5 6 7 8 9 10] ``` To learn more about this function, refer to the [official documentation](https://numpy.org/doc/stable/reference/generated/numpy.isfinite.html#numpy.isfinite) ## Remove Nan Values Using the `math.isnan` Method Apart from these two NumPy solutions, there are two more ways to remove `nan` values. These two ways involve `isnan()` function from `math` library and `isnull` function from `pandas` library. Both these functions check whether an element is `nan` or not and return a boolean result. Here is the solution using `isnan()` method. ``` import numpy as np import math myArray1 = np.array([1, 2, 3, np.nan, np.nan, 4, 5, 6, np.nan, 7, 8, 9, np.nan]) myArray2 = np.array([np.nan, np.nan, np.nan, np.nan, np.nan, np.nan]) myArray3 = np.array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10]) booleanArray1 = [not math.isnan(number) for number in myArray1] booleanArray2 = [not math.isnan(number) for number in myArray2] booleanArray3 = [not math.isnan(number) for number in myArray3] print(myArray1) print(myArray2) print(myArray3) print(myArray1[booleanArray1]) print(myArray2[booleanArray2]) print(myArray3[booleanArray3]) ``` Output: ``` [ 1. 2. 3. nan nan 4. 5. 6. nan 7. 8. 9. nan] [nan nan nan nan nan nan] [ 1 2 3 4 5 6 7 8 9 10] [1. 2. 3. 4. 5. 6. 7. 8. 9.] [] [ 1 2 3 4 5 6 7 8 9 10] ``` ## Remove Nan Values Using the `pandas.isnull` Method Below is the solution using the [`isnull()` method](https://www.delftstack.com/api/python-pandas/pandas-dataframe-dataframe.isnull-and-notnull-function/) from `pandas`. ``` import numpy as np import pandas as pd myArray1 = np.array([1, 2, 3, np.nan, np.nan, 4, 5, 6, np.nan, 7, 8, 9, np.nan]) myArray2 = np.array([np.nan, np.nan, np.nan, np.nan, np.nan, np.nan]) myArray3 = np.array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10]) booleanArray1 = [not pd.isnull(number) for number in myArray1] booleanArray2 = [not pd.isnull(number) for number in myArray2] booleanArray3 = [not pd.isnull(number) for number in myArray3] print(myArray1) print(myArray2) print(myArray3) print(myArray1[booleanArray1]) print(myArray2[booleanArray2]) print(myArray3[booleanArray3]) print(myArray1[~pd.isnull(myArray1)]) # Line 1 print(myArray2[~pd.isnull(myArray2)]) # Line 2 print(myArray3[~pd.isnull(myArray3)]) # Line 3 ``` Output: ``` [ 1. 2. 3. nan nan 4. 5. 6. nan 7. 8. 9. nan] [nan nan nan nan nan nan] [ 1 2 3 4 5 6 7 8 9 10] [1. 2. 3. 4. 5. 6. 7. 8. 9.] [] [ 1 2 3 4 5 6 7 8 9 10] [1. 2. 3. 4. 5. 6. 7. 8. 9.] [] [ 1 2 3 4 5 6 7 8 9 10] ``` Enjoying our tutorials? Subscribe to DelftStack on YouTube to support us in creating more high-quality video guides. [Subscribe](https://www.youtube.com/@delftstack/?sub_confirmation=1)