Finding optimal threshold based on ROC CURVE.
Project description
Installation
pip install threshold-finderUsage
Folowing display an usage example.
>>> from threshold_finder.finder import OptimalThresholdFinder, ThresholdFinder, YoudenThresholdFinder
>>> # Example data
>>> true_label = pd.Series([1,1,1,0,0,0])
>>> predicted_proba = pd.Series([0.9, 0.8, 0.7, 0.72, 0.6, 0.5])
>>> # Use a specific finder directly ...
>>> finder = YoudenThresholdFinder()
>>> optimal_threshold = finder.optimal_threshold(true_label, predicted_proba)
>>> print(optimal_threshold)
0.7
>>> # ... Or use the factory
>>> factory = ThresholdFinder()
>>> finder = factory.get_finder(method="youden_statistic")
>>> optimal_threshold = finder.optimal_threshold(true_label, predicted_proba)
>>> print(optimal_threshold)
0.7
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
File details
Details for the file threshold-finder-0.1.4.tar.gz.
File metadata
- Download URL: threshold-finder-0.1.4.tar.gz
- Upload date:
- Size: 1.9 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: poetry/1.1.13 CPython/3.8.10 Linux/5.13.0-1021-azure
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 | 94aad3133dd57bca8d8acd5759d6d3172d0c5cfc4831c884698d7427f58a052e |
|
| MD5 | b4524addcd76f1f0e93a2dac0f9bdca0 |
|
| BLAKE2b-256 | 2b29a2dde6925aae38a1d4e90e81c7b64cdbeae78fa08dbd625bce728a0dc6eb |
Provenance
File details
Details for the file threshold_finder-0.1.4-py3-none-any.whl.
File metadata
- Download URL: threshold_finder-0.1.4-py3-none-any.whl
- Upload date:
- Size: 2.2 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: poetry/1.1.13 CPython/3.8.10 Linux/5.13.0-1021-azure
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 | ef363552324e2e2cb0b0e55e30054978c344abdc8dc834bde01a7a1de962421a |
|
| MD5 | 49d82da42bf1fbd8a981b750e888b4fd |
|
| BLAKE2b-256 | 020eacfb1d5073e74309ea7969d532257140a0801301bfa9be3f50faab626813 |
