How to Calculate Precision and Recall in sklearn : Steps with expressions

Are you looking for How to calculate precision and recall sklearn ?  In this article, we will see the implementation of recall/precision. We will also explore the mathematical expression for precision and recall.

How to calculate precision and recall sklearn ?

Well! In order to give you a practice demonstration of precision recall implementation. We should build the model. But that will be completely out of the context here. Hence We will create dummy predicted array and dummy output array (real). Let’s see the implementation here.

1. Import the packages –

Here is the code for importing the packages.

import numpy as np
from sklearn.metrics import precision_recall_fscore_support

Here the NumPy package is for creating NumPy array. The other one sklearn.matrices package for the precision recall matrices.

2. dummy array creation (Optional) –

This is completely optional because in real scenarios we build the model. So we can skip this step.  Anyways here we create the dummy arrays.

y_true = np.array([0,1,1,0,1])
y_pred = np.array([0,1,0,0,1])

3. calculate precision and recall –

This is the final step, Here we will invoke the  precision_recall_fscore_support(). We will provide the above arrays in the above function.

precision_recall_fscore_support(y_true, y_pred, average='macro')

Here average is mainly for multiclass classification. The other two parameters are those dummy arrays. This function will return the f1_score also with the precision recall matrices.

Complete code –

If we combine the code from each section and merge at the place. Now, let’s run the code put with output.

Conclusion –

I hope this article must have explained the precision recall implementation using sklearn. Actually implementation wise It is a piece of cake. Well ! we are planning more content on precision recall like the theoretical section and use cases scenarios.  If you have any doubt over the same topic precision/recall, Please comment below in the section.

You may suggest more topic like this. We will surely provide you with the content. Please subscribe to us !!

Thanks

Data Science Learner Team