How to you find the cumulative sum in Python?

Understanding how to calculate cumulative sums is essential in various data analysis and statistics tasks. Whether you’re tracking the running total of values over time or need to create cumulative sum columns or rows in a dataset, the Numpy cumsum method provides a convenient and efficient way to achieve this.

python cumulative sum numpy cumsum

To calculate a cumulative sum in Python, you need to use the cumsum method from the NumPy library. Your array is the parameter which needs to be used. I created sample 3 x 3 array which I will sum up cumulatively.

import numpy as np

my_array = np.array([0, 20, 25, 40, 100, 300,
                     500, 800, 2000]).reshape(3, -1)
print(f"This is my array: \n {my_array}")
my_cumsum_array = np.cumsum(my_array)
print(f"This is the cumulative sum of my array"
      f":\n {my_cumsum_array}")

python cumulative sum numpy cumsum

As you can see Python summed every element up.

See also  Addressing ValueError: Resolving Shape Mismatch in NumPy Arrays

How to create a cumulative sum column in python?

It is also possible to calculate a cumulative sum column-wise by specifying an axis parameter in the cumsum method. You need to add axis parameters to cumsum method.

import numpy as np

my_array = np.array([0, 20, 25, 40, 100, 300,
                     500, 800, 2000]).reshape(3, -1)
print(f"This is my array: \n {my_array}")
my_cumsum_array = np.cumsum(my_array, axis=0)
print(f"This is the cumulative columns sum of my array"
      f":\n {my_cumsum_array}")

python cumulative column sum Numpy cumsum

Columns got sumed up.

See also  How to convert array from float to int in Numpy?

How to create a cumulative sum row in python?

Now, observe the difference when we change the axis value to 1 for row-wise cumulative sum calculation. This is how to calculate cum sum for a row.

import numpy as np

my_array = np.array([0, 20, 25, 40, 100, 300,
                     500, 800, 2000]).reshape(3, -1)
print(f"This is my array: \n {my_array}")
my_cumsum_array = np.cumsum(my_array, axis=1)
print(f"This is the cumulative rows sum of my array"
      f":\n {my_cumsum_array}")

python cumulative rows sum Numpy cumsum

And this is how rows got summed up.

See also  How to generate array filled with value in Numpy?