sklearn counterfactuals: Add option 'auto' for automatically choosing…

… a suitable optimization method - 'auto' is now the default in all high level API methods

ceml/sklearn/isolationforest.py

@@ -356,4 +356,7 @@ def isolationforest_generate_counterfactual(model, x, y_target, features_whiteli
     """
     cf = IsolationForestCounterfactual(model)
 
+    if optimizer == "auto":
+        optimizer = "nelder-mead"
+
     return cf.compute_counterfactual(x, y_target, features_whitelist, regularization, C, optimizer, return_as_dict)

ceml/sklearn/knn.py

@@ -212,4 +212,7 @@ def knn_generate_counterfactual(model, x, y_target, features_whitelist=None, dis
     """
     cf = KnnCounterfactual(model, dist)
 
+    if optimizer == "auto":
+        optimizer = "nelder-mead"
+
     return cf.compute_counterfactual(x, y_target, features_whitelist, regularization, C, optimizer, return_as_dict, done)

ceml/sklearn/lda.py

@@ -222,4 +222,7 @@ def lda_generate_counterfactual(model, x, y_target, features_whitelist=None, reg
     """
     cf = LdaCounterfactual(model)
 
+    if optimizer == "auto":
+        optimizer = "mp"
+
     return cf.compute_counterfactual(x, y_target, features_whitelist, regularization, C, optimizer, return_as_dict, done)

ceml/sklearn/linearregression.py

@@ -205,4 +205,7 @@ def linearregression_generate_counterfactual(model, x, y_target, features_whitel
     """
     cf = LinearRegressionCounterfactual(model)
 
+    if optimizer == "auto":
+        optimizer = "mp"
+
     return cf.compute_counterfactual(x, y_target, features_whitelist, regularization, C, optimizer, return_as_dict, done)

ceml/sklearn/models.py

@@ -15,7 +15,7 @@
 from .pipeline import pipeline_generate_counterfactual
 
 
-def generate_counterfactual(model, x, y_target, features_whitelist=None, dist="l2", regularization="l1", C=1.0, optimizer="nelder-mead", return_as_dict=True, done=None):
+def generate_counterfactual(model, x, y_target, features_whitelist=None, dist="l2", regularization="l1", C=1.0, optimizer="auto", return_as_dict=True, done=None):
     """Computes a counterfactual of a given input `x`.
 
     Parameters
@@ -72,9 +72,11 @@ def generate_counterfactual(model, x, y_target, features_whitelist=None, dist="l
         Name/Identifier of the optimizer that is used for computing the counterfactual.
         See :func:`ceml.optimizer.optimizer.desc_to_optim` for details.
 
+        Use "auto" if you do not know what optimizer to use - a suitable optimizer is chosen automatically.
+
         As an alternative, we can use any (custom) optimizer that is derived from the :class:`ceml.optim.optimizer.Optimizer` class.
 
-        The default is "nelder-mead".
+        The default is "auto".
     return_as_dict : `boolean`, optional
         If True, returns the counterfactual, its prediction and the needed changes to the input as dictionary.
         If False, the results are returned as a triple.

ceml/sklearn/pipeline.py

@@ -249,6 +249,8 @@ def pipeline_generate_counterfactual(model, x, y_target, features_whitelist=None
         Name/Identifier of the optimizer that is used for computing the counterfactual.
         See :func:`ceml.optim.optimizer.prepare_optim` for details.
 
+        Use "auto" if you do not know what optimizer to use - a suitable optimizer is chosen automatically.
+
         As an alternative, we can use any (custom) optimizer that is derived from the :class:`ceml.optim.optimizer.Optimizer` class.
 
         The default is "nelder-mead".
@@ -284,4 +286,7 @@ def pipeline_generate_counterfactual(model, x, y_target, features_whitelist=None
     """
     cf = PipelineCounterfactual(model)
 
+    if optimizer == "auto":
+        optimizer = "nelder-mead"
+
     return cf.compute_counterfactual(x, y_target, features_whitelist, regularization, C, optimizer, return_as_dict, done)

ceml/sklearn/randomforest.py

@@ -319,4 +319,7 @@ def randomforest_generate_counterfactual(model, x, y_target, features_whitelist=
     """
     cf = RandomForestCounterfactual(model)
 
+    if optimizer == "auto":
+        optimizer = "nelder-mead"
+
     return cf.compute_counterfactual(x, y_target, features_whitelist, regularization, C, optimizer, return_as_dict)

ceml/sklearn/softmaxregression.py

@@ -229,4 +229,7 @@ def softmaxregression_generate_counterfactual(model, x, y_target, features_white
     """
     cf = SoftmaxCounterfactual(model)
 
+    if optimizer == "auto":
+        optimizer = "mp"
+
     return cf.compute_counterfactual(x, y_target, features_whitelist, regularization, C, optimizer, return_as_dict, done)