`Numpy`

(abbreviation for ‘*Numerical Python*‘) is a library for performing large scale mathematical operations in fast and efficient manner. This article serves to educate you about methods one could use to iterate over columns in an 2D `NumPy`

array. Since a single dimensional array only consists of linear elements, there doesn’t exists a distinguished definition of rows and columns in them. Therefore, in order to perform such operations we need a array whose `len(ary.shape) > 1 `

.

To install `NumPy`

on your python environment, type the following code in your OS’s Command Processor (**CMD, Bash** etc):

pip install numpy

We would be taking a look at several methods of iterating over a column of an Array/Matrix:-

**METHOD 1:**

**CODE: Use of primitive 2D Slicing operation on an array to get the desired column/columns**

`import` `numpy as np ` ` ` `# Creating a sample numpy array (in 1D) ` `ary ` `=` `np.arange(` `1` `, ` `25` `, ` `1` `) ` ` ` `# Converting the 1 Dimensional array to a 2D array ` `# (to allow explicitly column and row operations) ` `ary ` `=` `ary.reshape(` `5` `, ` `5` `) ` ` ` `# Displaying the Matrix (use print(ary) in IDE) ` `print` `(ary) ` ` ` `# This for loop will iterate over all columns of the array one at a time ` `for` `col ` `in` `range` `(ary.shape[` `1` `]): ` ` ` `print` `(ary[:, col]) ` |

*chevron_right*

*filter_none*

**Output:**

[[ 0, 1, 2, 3, 4], [ 5, 6, 7, 8, 9], [10, 11, 12, 13, 14], [15, 16, 17, 18, 19], [20, 21, 22, 23, 24]]) [ 0 5 10 15 20] [ 1 6 11 16 21] [ 2 7 12 17 22] [ 3 8 13 18 23] [ 4 9 14 19 24]

**Explanation:**

In the above code, we firstly create an linear array of 25 elements (0-24) using * np.arange(25)*. Then we reshape (transform 1D to 2D) using

*to create a 2D array out of a linear array. Then we output the transformed array. Now we used a for loop which would iterate*

`np.reshape()`

*x*times (where x is the number of columns in the array) for which we used

*with the argument*

`range()`

*(where*

`ary.shape[1]`

*= number of columns in a 2D symmetric array). In each iteration we output a column out of the array using*

`shape[1]`

*which means that give give all elements of the column number =*

`ary[:, col]`

*.*

`col`

**METHOD 2:**

In this method we would transpose the array to treat each column element as a row element (which in turn is equivalent of column iteration).

**Code: **

`# libraries ` `import` `numpy as np ` ` ` `# Creating an 2D array of 25 elements ` `ary ` `=` `np.array([[ ` `0` `, ` `1` `, ` `2` `, ` `3` `, ` `4` `], ` ` ` `[ ` `5` `, ` `6` `, ` `7` `, ` `8` `, ` `9` `], ` ` ` `[` `10` `, ` `11` `, ` `12` `, ` `13` `, ` `14` `], ` ` ` `[` `15` `, ` `16` `, ` `17` `, ` `18` `, ` `19` `], ` ` ` `[` `20` `, ` `21` `, ` `22` `, ` `23` `, ` `24` `]]) ` ` ` ` ` `# This loop will iterate through each row of the transposed ` `# array (equivalent of iterating through each column) ` `for` `col ` `in` `ary.T: ` ` ` `print` `(col) ` |

*chevron_right*

*filter_none*

**Output:**

[ 0 5 10 15 20] [ 1 6 11 16 21] [ 2 7 12 17 22] [ 3 8 13 18 23] [ 4 9 14 19 24]

**Explanation:**

Firstly, we created an 2D array (same as the previous example) using * np.array()* and initialized it with 25 values. Then we transposed the array, using

*which in turn switches the rows with the columns and columns with the rows. Then we iterated over each row of this transposed array and printed the row values.*

`ary.T`

Attention geek! Strengthen your foundations with the **Python Programming Foundation** Course and learn the basics.

To begin with, your interview preparations Enhance your Data Structures concepts with the **Python DS** Course.