diff --git a/daal4py/.buildinfo b/daal4py/.buildinfo index 318f389..037945d 100755 --- a/daal4py/.buildinfo +++ b/daal4py/.buildinfo @@ -1,4 +1,4 @@ # Sphinx build info version 1 # This file hashes the configuration used when building these files. When it is not found, a full rebuild will be done. -config: b3f9dd5f9dffbf1afeba6d7177901aa9 +config: e1589d1aed29dd35620ebcc85aa22a53 tags: 645f666f9bcd5a90fca523b33c5a78b7 diff --git a/daal4py/.nojekyll b/daal4py/.nojekyll new file mode 100644 index 0000000..e69de29 diff --git a/daal4py/_modules/index.html b/daal4py/_modules/index.html index aac2e8a..eadb05b 100755 --- a/daal4py/_modules/index.html +++ b/daal4py/_modules/index.html @@ -1,15 +1,20 @@ - + - Overview: module code — daal4py 2021 documentation - - + Overview: module code — daal4py 2021.1 documentation + + + + + + + @@ -44,6 +49,9 @@ daal4py +
+ 2021 +
@@ -52,7 +60,7 @@

Conversion

@@ -149,7 +165,7 @@

Conversion

Classification and Regression Inference

The API is the same for classification and regression inference. -Based on the original model passed to the convert_model, d4p_prediction is either the classification or regression output.

+Based on the original model passed to the convert_model(), d4p_prediction is either the classification or regression output.

d4p_prediction = d4p_model.predict(test_data)
@@ -157,6 +173,35 @@

Classification and Regression Inferencepredict() method of d4p_model is being used to make predictions on the test_data dataset. The d4p_prediction variable stores the predictions made by the predict() method.

+

+
+

SHAP Value Calculation for Regression Models

+

SHAP contribution and interaction value calculation are natively supported by models created with daal4py Model Builders. +For these models, the predict() method takes additional keyword arguments:

+
+
d4p_model.predict(test_data, pred_contribs=True)      # for SHAP contributions
+d4p_model.predict(test_data, pred_interactions=True)  # for SHAP interactions
+
+
+
+

The returned prediction has the shape:

+
+
    +

  • (n_rows, n_features + 1) for SHAP contributions

    • +

  • (n_rows, n_features + 1, n_features + 1) for SHAP interactions

    • +
+
+

Here, n_rows is the number of rows (i.e., observations) in +test_data, and n_features is the number of features in the dataset.

+

The prediction result for SHAP contributions includes a feature attribution value for each feature and a bias term for each observation.

+

The prediction result for SHAP interactions comprises (n_features + 1) x (n_features + 1) values for all possible +feature combinations, along with their corresponding bias terms.

+
+

Note

+

The shapes of SHAP contributions and interactions are consistent with the XGBoost results. +In contrast, the SHAP Python package drops bias terms, resulting +in SHAP contributions (SHAP interactions) with one fewer column (one fewer column and row) per observation.

+

Scikit-learn-style Estimators

You can also use the scikit-learn-style classes GBTDAALClassifier and GBTDAALRegressor to convert and infer your models. For example:

@@ -168,13 +213,22 @@

Scikit-learn-style Estimators +

Limitations

+

Model Builders support only base inference with prediction and probabilities prediction. The functionality is to be extended. +Therefore, there are the following limitations: +- The categorical features are not supported for conversion and prediction. +- The multioutput models are not supported for conversion and prediction. +- SHAP values can be calculated for regression models only.

+

Examples

Model Builders models conversion

@@ -196,7 +250,7 @@

Articles and Blog Posts
-

© Copyright 2023, Intel.

+

© Copyright Intel.

Built with Sphinx using a diff --git a/daal4py/note.html b/daal4py/note.html deleted file mode 100644 index 91b715f..0000000 --- a/daal4py/note.html +++ /dev/null @@ -1,126 +0,0 @@ - - - - - - - <no title> — daal4py 2021 documentation - - - - - - - - - - - - - - - - -
- - -
- -
-
-
- -
-
-
-
- -
-

Note

-

Scikit-learn patching functionality in daal4py was deprecated and moved to a separate package, Intel(R) Extension for Scikit-learn*. -All future patches will be available only in Intel(R) Extension for Scikit-learn*. Use the scikit-learn-intelex package instead of daal4py for the scikit-learn acceleration.

-
- - -
-
- -
-
-
-
- - - - \ No newline at end of file diff --git a/daal4py/objects.inv b/daal4py/objects.inv index 712b18b..052223d 100755 Binary files a/daal4py/objects.inv and b/daal4py/objects.inv differ diff --git a/daal4py/scaling.html b/daal4py/scaling.html index 4670887..a06d5eb 100755 --- a/daal4py/scaling.html +++ b/daal4py/scaling.html @@ -1,16 +1,21 @@ - + - Scaling on Distributed Memory (Multiprocessing) — daal4py 2021 documentation - - + Scaling on Distributed Memory (Multiprocessing) — daal4py 2021.1 documentation + + + + + + + @@ -47,6 +52,9 @@ daal4py +
+ 2021 +
@@ -55,7 +63,7 @@
Built with Sphinx using a diff --git a/daal4py/search.html b/daal4py/search.html index aa3b8fc..06bb76c 100755 --- a/daal4py/search.html +++ b/daal4py/search.html @@ -1,16 +1,21 @@ - + - Search — daal4py 2021 documentation - - + Search — daal4py 2021.1 documentation + + + + + + + @@ -47,6 +52,9 @@ daal4py +
+ 2021 +
@@ -55,7 +63,7 @@
Built with Sphinx using a diff --git a/daal4py/streaming.html b/daal4py/streaming.html index 71a3447..1a253f7 100755 --- a/daal4py/streaming.html +++ b/daal4py/streaming.html @@ -1,16 +1,21 @@ - + - Streaming Data — daal4py 2021 documentation - - + Streaming Data — daal4py 2021.1 documentation + + + + + + + @@ -47,6 +52,9 @@ daal4py +
+ 2021 +
@@ -55,7 +63,7 @@
Built with Sphinx using a