gesund-val-library 0.1.0

Gesund.ai package for running validation metrics for classification, semantic segmentation, instance segmentation, and object detection models.

Validation Metrics Library

Overview

This library provides tools for calculating validation metrics for predictions and annotations in machine learning workflows. It includes a command-line tool for computing and displaying validation metrics.

Installation

To use this library, ensure you have the necessary dependencies installed in your environment. You can install them via pip:

pip install .

Usage

Command-Line Tool

The primary script for running validation metrics is run_metrics.py. This script calculates validation metrics based on JSON files containing predictions and annotations.

Arguments

• annotations (required): Path to the JSON file containing annotation data.
• predictions (required): Path to the JSON file containing prediction data.
• class_mappings (required): Path to the JSON file containing class_mappings data.
• --output (optional): Path to the file where the results will be saved. If not provided, the results will be printed to the console.

Example

1. Basic Usage: Print metrics to the console

   python run_metrics.py path/to/annotations.json path/to/predictions.json
   

2. Save Metrics to File: Save metrics to a specified file

   python -m scripts.run_metrics --annotations test_data/test_annotations_classification.json --predictions test_data/test_predictions_classification.json --class_mappings test_data/class_mappings.json --output path/to/output.json
   

This command will execute the metrics calculation and save the results to path/to/output.json. If the --output flag is not provided, the results will be displayed directly in the console.

Example JSON Inputs

• Annotations (test_annotations_classification.json):

  {
  "664df1bf782d9eb107789013": {
    "image_id": "664df1bf782d9eb107789013",
    "annotation": [
      {
        "id": "664dfb2085d8059c72b7b24a",
        "label": 0
      }
    ]
  },
  
  "664df1bf782d9eb107789015": {
    "image_id": "664df1bf782d9eb107789015",
    "annotation": [
      {
        "id": "664dfb2085d8059c72b7b24d",
        "label": 1
      }
    ]
  },
  ...
  }
  

• Predictions (test_predictions_classification.json):

  {
  "664df1bf782d9eb107789013": {
    "image_id": "664df1bf782d9eb107789013",
    "prediction_class": 1,
    "confidence": 0.731047693767988,
    "logits": [
      0.2689523062320121,
      0.731047693767988
    ],
    "loss": 16.11764907836914
  },
  
  "664df1bf782d9eb107789015": {
    "image_id": "664df1bf782d9eb107789015",
    "prediction_class": 1,
    "confidence": 0.7308736572776326,
    "logits": [
      0.26912634272236735,
      0.7308736572776326
    ],
    "loss": 0.007578411139547825
  },
  ...
  

• Class Mappings (test_class_mappings.json):

  {"0": "normal", "1": "pneumonia"}
  

Example Outputs

Console Output

When results are printed to the console, they will be in the following format:

Validation Metrics:
----------------------------------------
Accuracy:
    Validation: 0.4375
    Confidence_Interval: 0.2656 to 0.6094
----------------------------------------
Micro F1:
    Validation: 0.4375
    Confidence_Interval: 0.2656 to 0.6094
----------------------------------------
Macro F1:
    Validation: 0.4000
    Confidence_Interval: 0.2303 to 0.5697
----------------------------------------
AUC:
    Validation: 0.3996
    Confidence_Interval: 0.2299 to 0.5693
----------------------------------------
Precision:
    Validation: 0.4343
    Confidence_Interval: 0.2625 to 0.6060
----------------------------------------
Sensitivity:
    Validation: 0.4549
    Confidence_Interval: 0.2824 to 0.6274
----------------------------------------
Specificity:
    Validation: 0.4549
    Confidence_Interval: 0.2824 to 0.6274
----------------------------------------
Matthews C C:
    Validation: -0.1089
    Confidence_Interval: 0.0010 to 0.2168
----------------------------------------

Output File

If the --output flag is used, the metrics will be saved in the specified file path. The format of the saved file will be the same as the console output.