---
title: Recall
emoji: 🤗 
colorFrom: blue
colorTo: red
sdk: gradio
sdk_version: 3.19.1
app_file: app.py
pinned: false
tags:
- evaluate
- metric
description: >-
  Recall is the fraction of the positive examples that were correctly labeled by the model as positive. It can be computed with the equation:
  Recall = TP / (TP + FN)
  Where TP is the true positives and FN is the false negatives.
---

# Metric Card for Recall


## Metric Description

Recall is the fraction of the positive examples that were correctly labeled by the model as positive. It can be computed with the equation:
Recall = TP / (TP + FN)
Where TP is the number of true positives and FN is the number of false negatives.


## How to Use

At minimum, this metric takes as input two `list`s, each containing `int`s: predictions and references.

```python
>>> recall_metric = evaluate.load('recall')
>>> results = recall_metric.compute(references=[0, 1], predictions=[0, 1])
>>> print(results)
["{'recall': 1.0}"]
```


### Inputs
- **predictions** (`list` of `int`): The predicted labels.
- **references** (`list` of `int`): The ground truth labels.
- **labels** (`list` of `int`): The set of labels to include when `average` is not set to `binary`, and their order when average is `None`. Labels present in the data can be excluded in this input, for example to calculate a multiclass average ignoring a majority negative class, while labels not present in the data will result in 0 components in a macro average. For multilabel targets, labels are column indices. By default, all labels in y_true and y_pred are used in sorted order. Defaults to None.
- **pos_label** (`int`): The class label to use as the 'positive class' when calculating the recall. Defaults to `1`.
- **average** (`string`): This parameter is required for multiclass/multilabel targets. If None, the scores for each class are returned. Otherwise, this determines the type of averaging performed on the data. Defaults to `'binary'`.
    - `'binary'`: Only report results for the class specified by `pos_label`. This is applicable only if the target labels and predictions are binary.
    - `'micro'`: Calculate metrics globally by counting the total true positives, false negatives, and false positives.
    - `'macro'`: Calculate metrics for each label, and find their unweighted mean. This does not take label imbalance into account.
    - `'weighted'`: Calculate metrics for each label, and find their average weighted by support (the number of true instances for each label). This alters `'macro'` to account for label imbalance. Note that it can result in an F-score that is not between precision and recall.
    - `'samples'`: Calculate metrics for each instance, and find their average (only meaningful for multilabel classification).
- **sample_weight** (`list` of `float`): Sample weights Defaults to `None`.
- **zero_division** (): Sets the value to return when there is a zero division. Defaults to .
    - `'warn'`: If there is a zero division, the return value is `0`, but warnings are also raised.
    - `0`: If there is a zero division, the return value is `0`.
    - `1`: If there is a zero division, the return value is `1`.


### Output Values
- **recall**(`float`, or `array` of `float`, for multiclass targets): Either the general recall score, or the recall scores for individual classes, depending on the values input to `labels` and `average`. Minimum possible value is 0. Maximum possible value is 1. A higher recall means that more of the positive examples have been labeled correctly. Therefore, a higher recall is generally considered better.

Output Example(s):
```python
{'recall': 1.0}
```
```python
{'recall': array([1., 0., 0.])}
```

This metric outputs a dictionary with one entry, `'recall'`.


#### Values from Popular Papers


### Examples

Example 1-A simple example with some errors
```python
>>> recall_metric = evaluate.load('recall')
>>> results = recall_metric.compute(references=[0, 0, 1, 1, 1], predictions=[0, 1, 0, 1, 1])
>>> print(results)
{'recall': 0.6666666666666666}
```

Example 2-The same example as Example 1, but with `pos_label=0` instead of the default `pos_label=1`.
```python
>>> recall_metric = evaluate.load('recall')
>>> results = recall_metric.compute(references=[0, 0, 1, 1, 1], predictions=[0, 1, 0, 1, 1], pos_label=0)
>>> print(results)
{'recall': 0.5}
```

Example 3-The same example as Example 1, but with `sample_weight` included.
```python
>>> recall_metric = evaluate.load('recall')
>>> sample_weight = [0.9, 0.2, 0.9, 0.3, 0.8]
>>> results = recall_metric.compute(references=[0, 0, 1, 1, 1], predictions=[0, 1, 0, 1, 1], sample_weight=sample_weight)
>>> print(results)
{'recall': 0.55}
```

Example 4-A multiclass example, using different averages.
```python
>>> recall_metric = evaluate.load('recall')
>>> predictions = [0, 2, 1, 0, 0, 1]
>>> references = [0, 1, 2, 0, 1, 2]
>>> results = recall_metric.compute(predictions=predictions, references=references, average='macro')
>>> print(results)
{'recall': 0.3333333333333333}
>>> results = recall_metric.compute(predictions=predictions, references=references, average='micro')
>>> print(results)
{'recall': 0.3333333333333333}
>>> results = recall_metric.compute(predictions=predictions, references=references, average='weighted')
>>> print(results)
{'recall': 0.3333333333333333}
>>> results = recall_metric.compute(predictions=predictions, references=references, average=None)
>>> print(results)
{'recall': array([1., 0., 0.])}
```


## Limitations and Bias
1. Imbalanced Class Distribution - When the number of positive classes is much smaller than the negative classes, the recall may be inflated. This is because a model that predicts all cases as negative will have high accuracy but a low recall. Therefore it is important to consider other metrics like precision, F1-score, AUC-ROC to evaluate the performance of the models in imbalanced datasets.
2. Misclassification of negative classes - Recall only measures the ability of the model to identify positive cases correctly, without taking into account the misclassification of negative cases. Therefore, a model with a high recall may still misclassify a significant number of negative cases, which could have serious consequences depending on the application.
3. Bias towards models that predicts more positives - Recall is biased towards models that predict more positive cases, which may not be desirable in some cases. For example, in a medical diagnosis scenario, a model that predicts a high number of false positives may lead to unnecessary treatments or surgeries, causing harm to patients.
4. Dependency on the decision threshold - Recall, like other classification metrics, depends on the decision threshold used to classify cases into positive and negative classes. Depending on the problem, different decision thresholds may be more appropriate, and adjusting the threshold may affect the recall score.
5. Scenario/Context dependency - The importance of recall may vary depending on the context and the consequences of missing positive cases. For example, in a fraud detection system, missing a true positive case may have a higher cost than in other applications, such as email spam filtering.
6. Biases due to sampling - Recall can be biased by the way the data is sampled. For instance, if the dataset used for evaluation is not representative of the population to which the model will be applied, then the recall score may not generalize well to new data. Sampling bias can also arise when the positive cases are oversampled or undersampled, leading to an artificially high or low recall score.
7. Biases due to data quality - Recall can also be biased by the quality of the data. If the positive cases are noisy or mislabeled, then the recall score may be lower than expected. Similarly, if the negative cases are not well defined or ambiguous, then the recall score may be inflated.
8. Biases due to feature selection - Recall can be biased by the features used to train the model. If the features used to represent the positive cases are not sufficiently distinctive, then the recall score may be lower than expected. Conversely, if the features used to represent the negative cases overlap significantly with the positive cases, then the recall score may be inflated.
9. Biases due to Model Selection - Recall can also be biased by the model selection process. If the model is overfit to the training data, then the recall score may be artificially high. Conversely, if the model is underfit, then the recall score may be lower than expected.
10. Biases due to Model Bias - Recall can be biased by the way the labels are assigned. For example, if the labels are assigned by a human expert, then the recall score may be influenced by the expert's subjective judgment. Similarly, if the labels are assigned by an automated process, then the recall score may be influenced by the quality of the labeling algorithm.


## Citation(s)
```bibtex
@article{scikit-learn, title={Scikit-learn: Machine Learning in {P}ython}, author={Pedregosa, F. and Varoquaux, G. and Gramfort, A. and Michel, V. and Thirion, B. and Grisel, O. and Blondel, M. and Prettenhofer, P. and Weiss, R. and Dubourg, V. and Vanderplas, J. and Passos, A. and Cournapeau, D. and Brucher, M. and Perrot, M. and Duchesnay, E.}, journal={Journal of Machine Learning Research}, volume={12}, pages={2825--2830}, year={2011}
```


## Further References