Made eval_results for sklearn output the same structure as in the new training.py

Changed the name of eval_results to evals_result, so that the naming is the same in training.py and sklearn.py

Made the structure of evals_result the same as in training.py, the names of the keys are different:

In sklearn.py you cannot name your evals_result, but they are automatically called 'validation_0', 'validation_1' etc.
The dict evals_result will output something like: {'validation_0': {'logloss': ['0.674800', '0.657121']}, 'validation_1': {'logloss': ['0.63776', '0.58372']}}

In training.py you can name your multiple evals_result with a watchlist like: watchlist  = [(dtest,'eval'), (dtrain,'train')]
The dict evals_result will output something like: {'train': {'logloss': ['0.68495', '0.67691']}, 'eval': {'logloss': ['0.684877', '0.676767']}}

You can access the evals_result using the evals_result() function.

python-package/xgboost/sklearn.py

@@ -165,7 +165,7 @@ class XGBModel(XGBModelBase):
         """
         trainDmatrix = DMatrix(X, label=y, missing=self.missing)
 
-        eval_results = {}
+        evals_result = {}
         if eval_set is not None:
             evals = list(DMatrix(x[0], label=x[1]) for x in eval_set)
             evals = list(zip(evals, ["validation_{}".format(i) for i in
@@ -185,13 +185,14 @@ class XGBModel(XGBModelBase):
         self._Booster = train(params, trainDmatrix,
                               self.n_estimators, evals=evals,
                               early_stopping_rounds=early_stopping_rounds,
-                              evals_result=eval_results, feval=feval,
+                              evals_result=evals_result, feval=feval,
                               verbose_eval=verbose)
 
-        if eval_results:
+        if evals_result:
-            for val in eval_results.items():
+            for val in evals_result.items():
-                eval_results[val[0]] = [np.array(v[1], dtype=float) for v in val[1].items()]
+                evals_result_key = val[1].keys()[0]
-            self.eval_results = eval_results
+                evals_result[val[0]][evals_result_key] = val[1][evals_result_key]
+            self.evals_result_ = evals_result
 
         if early_stopping_rounds is not None:
             self.best_score = self._Booster.best_score
@@ -203,6 +204,41 @@ class XGBModel(XGBModelBase):
         test_dmatrix = DMatrix(data, missing=self.missing)
         return self.booster().predict(test_dmatrix)
     
+    def evals_result(self):
+        """Return the evaluation results.
+
+        If eval_set is passed to the `fit` function, you can call evals_result() to 
+        get evaluation results for all passed eval_sets. When eval_metric is also
+        passed to the `fit` function, the evals_result will contain the eval_metrics
+        passed to the `fit` function
+
+        Returns
+        -------
+        evals_result : dictionary
+        
+        Example
+        -------
+        param_dist = {'objective':'binary:logistic', 'n_estimators':2}
+        
+        clf = xgb.XGBModel(**param_dist)
+
+        clf.fit(X_train, y_train,
+                eval_set=[(X_train, y_train), (X_test, y_test)], 
+                eval_metric='logloss',
+                verbose=True)
+        
+        evals_result = clf.evals_result()
+        
+        The variable evals_result will contain: 
+        {'validation_0': {'logloss': ['0.604835', '0.531479']}, 'validation_1': {'logloss': ['0.41965', '0.17686']}}
+        """
+        if self.evals_result_:
+            evals_result = self.evals_result_
+        else:
+            raise Error('No results.')
+        
+        return evals_result
+
 
 class XGBClassifier(XGBModel, XGBClassifierBase):
     # pylint: disable=missing-docstring,too-many-arguments,invalid-name
@@ -259,7 +295,7 @@ class XGBClassifier(XGBModel, XGBClassifierBase):
             If `verbose` and an evaluation set is used, writes the evaluation
             metric measured on the validation set to stderr.
         """
-        eval_results = {}
+        evals_result = {}
         self.classes_ = list(np.unique(y))
         self.n_classes_ = len(self.classes_)
         if self.n_classes_ > 2:
@@ -299,13 +335,14 @@ class XGBClassifier(XGBModel, XGBClassifierBase):
         self._Booster = train(xgb_options, train_dmatrix, self.n_estimators,
                               evals=evals,
                               early_stopping_rounds=early_stopping_rounds,
-                              evals_result=eval_results, feval=feval,
+                              evals_result=evals_result, feval=feval,
                               verbose_eval=verbose)
 
-        if eval_results:
+        if evals_result:
-            for val in eval_results.items():
+            for val in evals_result.items():
-                eval_results[val[0]] = [np.array(v[1], dtype=float) for v in val[1].items()]
+                evals_result_key = val[1].keys()[0]
-            self.eval_results = eval_results
+                evals_result[val[0]][evals_result_key] = val[1][evals_result_key]
+            self.evals_result_ = evals_result
 
         if early_stopping_rounds is not None:
             self.best_score = self._Booster.best_score
@@ -333,6 +370,41 @@ class XGBClassifier(XGBModel, XGBClassifierBase):
             classzero_probs = 1.0 - classone_probs
             return np.vstack((classzero_probs, classone_probs)).transpose()
     
+    def evals_result(self):
+        """Return the evaluation results.
+
+        If eval_set is passed to the `fit` function, you can call evals_result() to 
+        get evaluation results for all passed eval_sets. When eval_metric is also
+        passed to the `fit` function, the evals_result will contain the eval_metrics
+        passed to the `fit` function
+
+        Returns
+        -------
+        evals_result : dictionary
+        
+        Example
+        -------
+        param_dist = {'objective':'binary:logistic', 'n_estimators':2}
+        
+        clf = xgb.XGBClassifier(**param_dist)
+
+        clf.fit(X_train, y_train,
+                eval_set=[(X_train, y_train), (X_test, y_test)], 
+                eval_metric='logloss',
+                verbose=True)
+        
+        evals_result = clf.evals_result()
+        
+        The variable evals_result will contain: 
+        {'validation_0': {'logloss': ['0.604835', '0.531479']}, 'validation_1': {'logloss': ['0.41965', '0.17686']}}
+        """
+        if self.evals_result_:
+            evals_result = self.evals_result_
+        else:
+            raise Error('No results.')
+        
+        return evals_result
+
 class XGBRegressor(XGBModel, XGBRegressorBase):
     # pylint: disable=missing-docstring
     __doc__ = """Implementation of the scikit-learn API for XGBoost regression.