In this tutorial, we will look at how to sort a numpy array by a column with the help of some examples.
To sort a numpy array by a given column, use the following steps –
- Get the column index, i. For example, if you want to sort the array by values in the second column, i will be 1 (rows and columns and indexed starting from 0).
- Get the index array of the rows after sorting by the above column using the
argsort()
method. - Use the above index array to reorder the rows.
The following is the syntax –
# sort 2d numpy array by the column with index i sorted_arr = arr[arr[:, i].argsort()]
The above code sorts the array by the values in the column with the index i
in ascending order. To sort in descending order, reverse the values resulting from the argsort()
function.

Let’s now look at some examples of using the above syntax –
Example 1 – Sort 2d array by a column in ascending order
import numpy as np # create a numpy array arr = np.array([ [1, 7, 1], [2, 5, 0], [3, 9, 0] ]) # sort the above array by the second column, i=1 sorted_arr = sorted_arr = arr[arr[:, 1].argsort()] print(sorted_arr)
Output:
[[2 5 0] [1 7 1] [3 9 0]]
You can see that the resulting array is sorted by the second column in ascending order.
Example 2 – Sort 2d array by a column in descending order
We can similarly sort a 2d array (by a specific column) in descending order. To do this, reverse the row indices obtained from the argsort()
method. This will give the array of row indices when sorted in descending order.
Let’s take the same example as above.
import numpy as np # create a numpy array arr = np.array([ [1, 7, 1], [2, 5, 0], [3, 9, 0] ]) # sort the above array in descending order by the second column, i=1 sorted_arr = sorted_arr = arr[arr[:, 1].argsort()[::-1]] print(sorted_arr)
Output:
[[3 9 0] [1 7 1] [2 5 0]]
The resulting array is sorted in descending order by the second column.
You might also be interested in –
Subscribe to our newsletter for more informative guides and tutorials.
We do not spam and you can opt out any time.