In this article, we will discuss how to find index of a value in a NumPy array (both 1D & 2D) using numpy.where().

Let’s create a Numpy array from a list of numbers i.e.

import numpy as np # Create a numpy array from a list of numbers arr = np.array([11, 12, 13, 14, 15, 16, 17, 15, 11, 12, 14, 15, 16, 17])

Now let’s see how to to search elements in this NumPy array.

## Find index of a value in 1D Numpy array

In the above numpy array, elements with value **15** occurs at different places let’s find all it’s indices i.e.

# Get the index of elements with value 15 result = np.where(arr == 15) print('Tuple of arrays returned : ', result) print("Elements with value 15 exists at following indices", result[0], sep='\n')

**Output:**

Tuple of arrays returned : (array([ 4, 7, 11], dtype=int32),) Elements with value 15 exists at following indices [ 4 7 11]

**result** is a tuple of arrays (one for each axis) containing the indices where value 15 exists in array arr i.e.

(array([ 4, 7, 11], dtype=int32),)

As our array **arr** is a flat 1D array, so returned tuple will contain only one array of indices and contents of the returned array **result[0]** are,

### Frequently Asked:

- Create Numpy Array of different shapes & initialize with identical values using numpy.full() in Python
- Remove Last N Elements from a NumPy Array
- What is a Structured Numpy Array and how to create and sort it in Python?
- Normalize a NumPy array to a unit vector in Python

[ 4 7 11]

Get the first index of element with value 15,

result[0][0]

**How did it worked ?**

numpy.where() accepts a condition and 2 optional arrays i.e.

numpy.where(condition[, x, y])

If only condition argument is given then it returns the indices of the elements which are TRUE in bool numpy array returned by condition. For example following condition,

boolArr = (arr == 15)

returns a bool numpy array **boolArr**, containing TRUE of each element that is equal to 15, for other elements it contains False i.e.

[False False False False True False False True False False False True False False]

Now if you will pass this bool numpy array to numpy.where()

result = numpy.where(boolArr)

Then it will return a tuple of arrays (one for each axis) containing indices where value was TRUE in given bool numpy array i.e.

[ 4 7 11]

### If element not found in numpy array

If the given element doesn’t exist in numpy array then returned array of indices will be empty i.e.

# If given element doesn't exist in the array then it will return an empty array result = np.where(arr == 111) print('Empty Array returned : ', result) print("value 111 exists at following indices", result[0], sep='\n')

**Output:**

Empty Array returned : (array([], dtype=int32),) value 111 exists at following indices []

## Find index of a value in 2D Numpy array | Matrix

Let’s create a 2D numpy array i.e.

# Create a 2D Numpy array from list of lists arr = np.array([[11, 12, 13], [14, 15, 16], [17, 15, 11], [12, 14, 15]])

Contents of the 2D numpy array are,

[[11 12 13] [14 15 16] [17 15 11] [12 14 15]]

Let’s find the indices of element with **value 15** in this 2D numpy array i.e.

# Get the index of elements with value 15 result = np.where(arr == 15) print('Tuple of arrays returned : ', result)

Output:

Tuple of arrays returned : (array([1, 2, 3], dtype=int32), array([1, 1, 2], dtype=int32))

It returns a tuple of arrays one for each dimension. Like in our case it’s a two dimension array, so numpy.where() will returns a tuple of two arrays.

Now returned **array 1** represents the **row indices** where this value is found i.e.

[1, 2, 3]

Whereas, **array 2** represents the **column indices** where this value is found i.e.

[1, 1, 2]

Length of both the arrays will be same. So to get the list of exact coordinates we can zip these arrays i.e.

# zip the 2 arrays to get the exact coordinates listOfCoordinates= list(zip(result[0], result[1]))

Now let’s iterate over the list of coordinates and print them i.e.

# iterate over the list of coordinates for cord in listOfCoordinates: print(cord)

Coordinates of 2d Numpy array where element with value exist i.e.

(1, 1) (2, 1) (3, 2)

## Get indices of elements based on multiple conditions

When can also pass multiple conditions to numpy.where(). For example, get the indices of elements with value less than 16 and greater than 12 i.e.

# Create a numpy array from a list of numbers arr = np.array([11, 12, 13, 14, 15, 16, 17, 15, 11, 12, 14, 15, 16, 17]) # Get the index of elements with value less than 16 and greater than 12 result = np.where((arr > 12) & (arr < 16)) print("Elements with value less than 16 and greater than 12 exists at following indices", result, sep='\n')

Output:

Elements with value less than 16 and greater than 12 exists at following indices (array([ 2, 3, 4, 7, 10, 11], dtype=int32),)

## Get the first index of an element in numpy array

result = np.where(arr == 15) if len(result) > 0 and len(result[0]) > 0: print('First Index of element with value 15 is ', result[0][0])

**Output**

First Index of element with value 15 is 4

**Complete example is as follows,**

import numpy as np print("*** Find the index of an element in 1D Numpy Array ***") # Create a numpy array from a list of numbers arr = np.array([11, 12, 13, 14, 15, 16, 17, 15, 11, 12, 14, 15, 16, 17]) # Get the index of elements with value 15 result = np.where(arr == 15) print('Tuple of arrays returned : ', result) print("Elements with value 15 exists at following indices", result[0], sep='\n') print('First Index of element with value 15 is : ', result[0][0]) # If given element doesn't exist in the array then it will return an empty array result = np.where(arr == 111) print('Empty Array returned : ', result) print("value 111 exists at following indices", result[0], sep='\n') print("*** Find the index of an element in 2D Numpy Array ***") # Create a 2D Numpy array from list of lists arr = np.array([[11, 12, 13], [14, 15, 16], [17, 15, 11], [12, 14, 15]]) print('Contents of 2D Numpy Array', arr, sep='\n') # Get the index of elements with value 17 result = np.where(arr == 15) print('Tuple of arrays returned : ', result) print('List of coordinates where element with value 15 exists in given 2D array : ') # zip the 2 arrays to get the exact coordinates listOfCoordinates = list(zip(result[0], result[1])) # iterate over the list of coordinates for cord in listOfCoordinates: print(cord) print("*** Get the index of an element based on multiple conditions Numpy Array ***") # Create a numpy array from a list of numbers arr = np.array([11, 12, 13, 14, 15, 16, 17, 15, 11, 12, 14, 15, 16, 17]) # Get the index of elements with value less than 16 and greater than 12 result = np.where((arr > 12) & (arr < 16)) print("Elements with value less than 16 and greater than 12 exists at following indices", result, sep='\n') print("*** Get the first index of an element in Numpy Array ***") result = np.where(arr == 15) if len(result) > 0 and len(result[0]) > 0: print('First Index of element with value 15 is ', result[0][0])

**Output:**

*** Find the index of an element in 1D Numpy Array *** Tuple of arrays returned : (array([ 4, 7, 11], dtype=int32),) Elements with value 15 exists at following indices [ 4 7 11] First Index of element with value 15 is : 4 Empty Array returned : (array([], dtype=int32),) value 111 exists at following indices [] *** Find the index of an element in 2D Numpy Array *** Contents of 2D Numpy Array [[11 12 13] [14 15 16] [17 15 11] [12 14 15]] Tuple of arrays returned : (array([1, 2, 3], dtype=int32), array([1, 1, 2], dtype=int32)) List of coordinates where element with value 15 exists in given 2D array : (1, 1) (2, 1) (3, 2) *** Get the index of an element based on multiple conditions Numpy Array *** Elements with value less than 16 and greater than 12 exists at following indices (array([ 2, 3, 4, 7, 10, 11], dtype=int32),) *** Get the first index of an element in Numpy Array *** First Index of element with value 15 is 4

## Summary

Today, we learned about different ways to find the indices of elements in a NumPy Array.

JimJust wanted to say this page was EXTREMELY helpful for me. I was stuck on a problem for hours and then found exactly what I was looking for here (info about np.where and 2D matrices). Thanks so much!!