How Can I Sort a NumPy Array by a Specific Column?

Sorting a NumPy Array by Column

Sorting a NumPy array by its specific column can often be a necessary operation for data manipulation. For instance, consider the array:

a = array([[9, 2, 3],
           [4, 5, 6],
           [7, 0, 5]])

Suppose we want to reorder the rows of a based on the values in the second column. To achieve this, we can utilize the argsort() function in NumPy.

a[:, 1].argsort()

The above expression applies argsort() to the second column of a, producing an array of indices that correspond to the sorted values. Specifically, it returns:

array([1, 0, 2])

These indices represent the order in which the rows of a should be reordered to sort by the second column in ascending order. Combining this with array indexing, we can obtain the sorted array:

a[a[:, 1].argsort()]

This expression returns:

array([[7, 0, 5],
       [9, 2, 3],
       [4, 5, 6]])

As desired, the rows of a have been sorted according to the second column.

The above is the detailed content of How Can I Sort a NumPy Array by a Specific Column?. For more information, please follow other related articles on the PHP Chinese website!