https://github.com/guillaumegenthial/tf_metrics

Multi-class metrics for Tensorflow
https://github.com/guillaumegenthial/tf_metrics

f1 metrics multiclass precision recall tensorflow tensorflow-estimator

Last synced: 8 days ago
JSON representation

Multi-class metrics for Tensorflow

• Host: GitHub
• URL: https://github.com/guillaumegenthial/tf_metrics
• Owner: guillaumegenthial
• License: apache-2.0
• Created: 2018-09-16T17:37:21.000Z (about 6 years ago)
• Default Branch: master
• Last Pushed: 2022-09-20T09:33:09.000Z (about 2 years ago)
• Last Synced: 2023-11-07T18:17:56.111Z (about 1 year ago)
• Topics: f1, metrics, multiclass, precision, recall, tensorflow, tensorflow-estimator
• Language: Python
• Size: 15.6 KB
• Stars: 223
• Watchers: 10
• Forks: 72
• Open Issues: 6
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# TF Metrics

[![Build Status](https://travis-ci.org/guillaumegenthial/tf_metrics.svg?branch=master)](https://travis-ci.org/guillaumegenthial/tf_metrics)

Multi-class metrics for Tensorflow, similar to scikit-learn multi-class metrics.

*Thank you all for making this project live (50-100 clones/day 😎). Contributions welcome!*

## Install

To add `tf_metrics` to your current python environment, run

```

pip install git+https://github.com/guillaumegenthial/tf_metrics.git

```

For a more advanced use (editable mode, for developers)

```

git clone https://github.com/guillaumegenthial/tf_metrics.git

cd tf_metrics

pip install -r requirements.txt

```

## Example

Pre-requisite: understand the general `tf.metrics` API. See for instance [the official guide on custom estimators](https://www.tensorflow.org/guide/custom_estimators#evaluate) or the [official documentation](https://www.tensorflow.org/api_docs/python/tf/metrics/accuracy).

Simple example

```python

import tensorflow as tf

import tf_metrics

y_true = [0, 1, 0, 0, 0, 2, 3, 0, 0, 1]

y_pred = [0, 1, 0, 0, 1, 2, 0, 3, 3, 1]

pos_indices = [1, 2, 3]  # Class 0 is the 'negative' class

num_classes = 4

average = 'micro'

# Tuple of (value, update_op)

precision = tf_metrics.precision(

    y_true, y_pred, num_classes, pos_indices, average=average)

recall = tf_metrics.recall(

    y_true, y_pred, num_classes, pos_indices, average=average)

f2 = tf_metrics.fbeta(

    y_true, y_pred, num_classes, pos_indices, average=average, beta=2)

f1 = tf_metrics.f1(

    y_true, y_pred, num_classes, pos_indices, average=average)

# Run the update op and get the updated value

with tf.Session() as sess:

    sess.run(tf.local_variables_initializer())

    sess.run(precision[1])

```

If you want to use it with `tf.estimator.Estimator`, add to your `model_fn`

```python

metrics = {

    'precision': precision,

    'recall': recall,

    'f1': f1,

    'f2': f2

    }

# For Tensorboard

for metric_name, metric in metrics.items():

    tf.summary.scalar(metric_name, metric[1])

if mode == tf.estimator.ModeKeys.EVAL:

    return tf.estimator.EstimatorSpec(

        mode, loss=loss, eval_metric_ops=metrics)

```