Hello everyone. Today I am again with a new post regarding mathematical and arithmetic operations using NumPy package. The last post deals with manipulating an array in NumPy. The link to the post along with other previous posts can be found at the bottom of this blog.
First of all we will start by creating three arrays: one array is one dimensional and the rest two is two dimensional of shape 53 and 35 respectively. We will then print the content of each of these arrays.
import numpy as np
array_1 = np.array([1,3,6,8,9])
array_2 = np.array([2,3,4,2,1,9,7,5,8,6,1,1,9,2,3]).reshape(5,3)
array_3 = np.array([2,3,4,2,1,9,7,5,8,6,1,1,9,2,3]).reshape(3,5)
array_1
array_2
array_3
The output is:
First lets see the addition of array. We will try to add array_1 with itself. There are two ways to do it: the first is using + operator and the other way is using add() function.
array_1 + array_1
np.add(array_1, array_1)
You can see the output as:
Now lets try adding array_2 and array_3. Remember the shape of first one is 53 and the second one is 35. If we try to add these two arrays, we will get following error. The reason is that arrays have different shape along two dimensions that make broadcasting impossible.
Lets try adding array_3 with array_1.
array_1
array_3
array_3 + array_1
Now we can see the output as:
In the above program, the smaller array is being broadcasted to the bigger array to make the shape compatible. In the same way broadcasting works for multiplication as well:
array_3 * 2
array_3 * array_1
In the same way we can divide the array and do other operation like subtraction. Now lets see other things like getting the sum using sum() function.
array_3
np.sum(array_3)
np.sum(array_3, axis=0)
np.sum(array_3, axis=1)
If you see the output, you will notice the sum function returns the total value of all elements inside the array. Specifying the axis will return the value either across rows or columns. axis=0 does operation column-wise and axis=1 does operation along the rows.
There is another interesting function called cumsum() that helps to find the cumulative sum. Again you can specify the axis of your choice.
array_3
np.cumsum(array_3)
np.cumsum(array_3, axis=0)
The output is:
Next we will see various functions like mean, minimum and maximum value.
array_2
array_2.mean()
np.mean(array_2, axis=1)
np.mean(array_2, axis=0)
np.amax(array_2)
np.amin(array_2)
In the above code, we tried finding the mean of all element of array_2, along the rows and column, and the maximum and minimum value of element inside it.
Link to previous NumPy Post: