Implementations for KDEy Methods

docs/source/api.md

@@ -1,11 +1,16 @@
 # API
 
-The `GenericMethod` defines the interface for many common quantification and unfolding algorithms. Most importantly, this interface consists of their `fit` and `predict` methods.
+The `AbstractMethod` defines the interface for all quantification and unfolding algorithms. Most importantly, this interface consists of their `fit` and `predict` methods.
 
-Instances of [](#popular-algorithms) for quantification and unfolding are created through the corresponding constructors. However, you can also define your own quantification methods as a `GenericMethod` that combines an arbitrary choice of [](#losses), [](#regularizers) and [](#feature-transformations).
+```{eval-rst}
+.. autoclass:: qunfold.AbstractMethod
+   :members:
+```
+
+Instances of [](#popular-algorithms) for quantification and unfolding are created through the corresponding constructors. However, you can also define your own quantification methods as a `LinearMethod` that combines an arbitrary choice of [](#losses), [](#regularizers) and [](#feature-transformations).
 
 ```{eval-rst}
-.. autoclass:: qunfold.GenericMethod
+.. autoclass:: qunfold.LinearMethod
    :members:
 ```
 

docs/source/developer-guide.md

@@ -36,12 +36,12 @@ As soon as you push to the `main` branch, GitHub Actions will build the document
 
 ## Custom implementations
 
-Custom [](#losses) and [](#feature-transformations) can be used in any instance of `GenericMethod`. Use the already existing implementations as examples.
+Custom [](#losses) and [](#feature-transformations) can be used in any instance of `LinearMethod`. Use the already existing implementations as examples.
 
 
 ### Losses
 
-The most convenient way of implementing a custom loss is to create a [JAX](https://jax.readthedocs.io/)-powered function `(p, q, M, N) -> loss_value`. From this function, you can create a *FunctionLoss* object to be used in any instance of `GenericMethod`.
+The most convenient way of implementing a custom loss is to create a [JAX](https://jax.readthedocs.io/)-powered function `(p, q, M, N) -> loss_value`. From this function, you can create a *FunctionLoss* object to be used in any instance of `LinearMethod`.
 
 ```{eval-rst}
 .. autoclass:: qunfold.losses.FunctionLoss

docs/source/index.md

@@ -66,5 +66,5 @@ loss = TikhonovRegularized(LeastSquaresLoss(), 0.01)
 transformer = ClassTransformer(CVClassifier(LogisticRegression(), 10))
 
 # the ordinal variant of ACC, ready for being used in QuaPy
-ordinal_acc = QuaPyWrapper(GenericMethod(loss, transformer))
+ordinal_acc = QuaPyWrapper(LinearMethod(loss, transformer))
 ```

qunfold/__init__.py

@@ -7,6 +7,9 @@
     CombinedLoss,
     TikhonovRegularization,
     TikhonovRegularized,
+    KDEyHDLoss,
+    KDEyCSLoss,
+    KDEyMLLoss,
 )
 
 from .transformers import (
@@ -18,10 +21,14 @@
     LaplacianKernelTransformer,
     GaussianKernelTransformer,
     GaussianRFFKernelTransformer,
+    KDEyHDTransformer,
+    KDEyCSTransformer,
+    KDEyMLTransformerID,
 )
 
-from .methods import (
-    GenericMethod,
+from .methods.linear import (
+    AbstractMethod,
+    LinearMethod,
     ACC,
     PACC,
     RUN,
@@ -30,4 +37,12 @@
     EDx,
     EDy,
     KMM,
+    KDEyHD,
+    KDEyCS,
+    KDEyMLID,
+)
+
+from .methods.kernel_density import (
+    KDEyML,
+    KDEyMLQP,
 )