Skip to main content

Add-on containing scoring engine widgets for PFA, PMML and ONNX (coming soon) models

Project description

Orange3 Scoring

This is an scoring/inference add-on for Orange3. This add-on adds widgets to load PMML and PFA models and score data.

Dependencies

To use PMML models make sure you have Java installed:

  • Java >= 1.8
  • pypmml (downloaded during installation)

To use PFA models:

  • titus2 (downloaded during installation)

Installation

To install the add-on using pip, run

pip install orange3-scoring

To register this add-on with Orange, but keep the code in the development directory (do not copy it to Python's site-packages directory), run

pip install -e .

Issues, Questions and Feature Requests

Please raise an issue/question/request here.

Development

Want to contribute? Great!

Please raise an issue to discuss your ideas and send a pull request.

Usage

After the installation, the widget from this add-on is registered with Orange. To run Orange from the terminal, use

python -m Orange.canvas

The new widget appears in the toolbox bar under the section Scoring.

01_intro

Drag and drop the Load PMML/PFA Model widget.

02_loadmodel

Load your PMML model and inspect Input and Output field(s). Sample PMML File here.

03_loadmodel_pmml

Add input dataset using File widget (iris) and connect the two widgets to Evaluate PMML/PFA Model widget. You can inspect the fields in data and the model and view Processing INFO or Errors.

04_evaluate_load

Now hit Score button to score.

05_evaluate_score

Connect the output to Data Table widget to view the results. 3 new columns (cluster, cluster_name & distance) are added after scoring the data obtained for each input record. The actual class value present in the data is also converted to metadata of the result table.

06_view_result

Now lets load a PFA Model. Sample PFA File here.

07_loadmodel_pfa

Score the data using new PFA Model.

08_evaluate_load

Now hit Score button to score.

09_evaluate_score

View the results. You can see the predicted class for iris as provided by the PFA Model.

10_view_result

Another output signal is produced which contains the Evaluation Results which can be connected to Confusion Matrix, ROC Analysis and Lift Curve widgets. We can connect it to the Confusion Matrix widget to view the difference in predicted and actual results.

11_view_confusion

Release history Release notifications

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for orange3-scoring, version 0.0.1
Filename, size File type Python version Upload date Hashes
Filename, size orange3_scoring-0.0.1-py3-none-any.whl (28.3 kB) File type Wheel Python version py3 Upload date Hashes View hashes
Filename, size orange3-scoring-0.0.1.tar.gz (21.2 kB) File type Source Python version None Upload date Hashes View hashes

Supported by

Elastic Elastic Search Pingdom Pingdom Monitoring Google Google BigQuery Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN DigiCert DigiCert EV certificate StatusPage StatusPage Status page