[python] [docs] fixed objective in sklearn wrapper; added missed obje…

…ctives & metrics to docs (#1059)

* added missed aliases for task parameter

* fixed indents

* added missed aliases and options for tree_learner parameter

* added missed objectives to docs

* fixed typo in Poisson parameter and its description

* fixed model_format parameter description

* added missed metrics to docs

* fixed sklearn objective

* fixed set_params

* fixed docs

* added missed options to objectives

* added note about ignore_column (#1061)

docs/Parameters.rst

@@ -39,22 +39,22 @@ Core Parameters
 
    -  path of config file
 
--  ``task``, default=\ ``train``, type=enum, options=\ ``train``, ``prediction``
+-  ``task``, default=\ ``train``, type=enum, options=\ ``train``, ``predict``, ``convert_model``
 
-   -  ``train`` for training
+   -  ``train``, alias=\ ``training``, for training
 
-   -  ``prediction`` for prediction.
+   -  ``predict``, alias=\ ``prediction``, ``test``, for prediction.
 
-   -  ``convert_model`` for converting model file into if-else format, see more information in `Convert model parameters <#convert-model-parameters>`__
+   -  ``convert_model``, for converting model file into if-else format, see more information in `Convert model parameters <#convert-model-parameters>`__
 
 -  ``application``, default=\ ``regression``, type=enum,
-   options=\ ``regression``, ``regression_l2``, ``regression_l1``, ``huber``, ``fair``, ``poisson``, ``quantile``, ``quantile_l2``,
-   ``binary``, ``lambdarank``, ``multiclass``,
+   options=\ ``regression``, ``regression_l1``, ``huber``, ``fair``, ``poisson``, ``quantile``, ``quantile_l2``,
+   ``binary``, ``multiclass``, ``multiclassova``, ``xentropy``, ``xentlambda``, ``lambdarank``,
    alias=\ ``objective``, ``app``
 
-   -  ``regression``, regression application
+   -  regression application
 
-      -  ``regression_l2``, L2 loss, alias=\ ``mean_squared_error``, ``mse``
+      -  ``regression_l2``, L2 loss, alias=\ ``regression``, ``mean_squared_error``, ``mse``
 
       -  ``regression_l1``, L1 loss, alias=\ ``mean_absolute_error``, ``mae``
 
@@ -68,16 +68,28 @@ Core Parameters
 
       -  ``quantile_l2``, like the ``quantile``, but L2 loss is used instead
 
-   -  ``binary``, binary classification application
+   -  ``binary``, binary `log loss`_ classification application
+
+   -  multi-class classification application
+
+      -  ``multiclass``, `softmax`_ objective function, ``num_class`` should be set as well
+
+      -  ``multiclassova``, `One-vs-All`_ binary objective function, ``num_class`` should be set as well
+
+   -  cross-entropy application
+
+      -  ``xentropy``, objective function for cross-entropy (with optional linear weights), alias=\ ``cross_entropy``
+
+      -  ``xentlambda``, alternative parameterization of cross-entropy, alias=\ ``cross_entropy_lambda``
+
+      -  the label is anything in interval [0, 1]
 
    -  ``lambdarank``, `lambdarank`_ application
 
       -  the label should be ``int`` type in lambdarank tasks, and larger number represent the higher relevance (e.g. 0:bad, 1:fair, 2:good, 3:perfect)
 
       -  ``label_gain`` can be used to set the gain(weight) of ``int`` label
 
-   -  ``multiclass``, multi-class classification application, ``num_class`` should be set as well
-
 -  ``boosting``, default=\ ``gbdt``, type=enum,
    options=\ ``gbdt``, ``rf``, ``dart``, ``goss``,
    alias=\ ``boost``, ``boosting_type``
@@ -120,13 +132,15 @@ Core Parameters
 
    -  number of leaves in one tree
 
--  ``tree_learner``, default=\ ``serial``, type=enum, options=\ ``serial``, ``feature``, ``data``, alias=\ ``tree``
+-  ``tree_learner``, default=\ ``serial``, type=enum, options=\ ``serial``, ``feature``, ``data``, ``voting``, alias=\ ``tree``
 
    -  ``serial``, single machine tree learner
 
-   -  ``feature``, feature parallel tree learner
+   -  ``feature``, alias=\ ``feature_parallel``, feature parallel tree learner
 
-   -  ``data``, data parallel tree learner
+   -  ``data``, alias=\ ``data_parallel``, data parallel tree learner
+
+   -  ``voting``, alias=\ ``voting_parallel``, voting parallel tree learner
 
    -  refer to `Parallel Learning Guide <./Parallel-Learning-Guide.rst>`__ to get more details
 
@@ -321,7 +335,7 @@ IO Parameters
 
    -  file name of prediction result in ``prediction`` task
 
--  ``model_format``, default=\ ``text``, type=string
+-  ``model_format``, default=\ ``text``, type=multi-enum, options=\ ``text``, ``proto``
 
    -  format to save and load model
 
@@ -406,6 +420,8 @@ IO Parameters
 
    -  add a prefix ``name:`` for column name, e.g. ``ignore_column=name:c1,c2,c3`` means c1, c2 and c3 will be ignored
 
+   -  **Note**: works only in CLI-version
+
    -  **Note**: index starts from ``0``. And it doesn't count the label column
 
 -  ``categorical_feature``, default=\ ``""``, type=string, alias=\ ``categorical_column``, ``cat_feature``, ``cat_column``
@@ -507,9 +523,9 @@ Objective Parameters
 
    -  parameter to control the width of Gaussian function. Will be used in ``regression_l1`` and ``huber`` losses
 
--  ``poission_max_delta_step``, default=\ ``0.7``, type=double
+-  ``poisson_max_delta_step``, default=\ ``0.7``, type=double
 
-   -  parameter used to safeguard optimization
+   -  parameter for `Poisson regression`_ to safeguard optimization
 
 -  ``scale_pos_weight``, default=\ ``1.0``, type=double
 
@@ -579,13 +595,18 @@ Metric Parameters
 
    -  ``binary_logloss``, `log loss`_
 
-   -  ``binary_error``.
-      For one sample: ``0`` for correct classification, ``1`` for error classification
+   -  ``binary_error``, for one sample: ``0`` for correct classification, ``1`` for error classification
 
    -  ``multi_logloss``, log loss for mulit-class classification
 
    -  ``multi_error``, error rate for mulit-class classification
 
+   -  ``xentropy``, cross-entropy (with optional linear weights), alias=\ ``cross_entropy``
+
+   -  ``xentlambda``, "intensity-weighted" cross-entropy, alias=\ ``cross_entropy_lambda``
+
+   -  ``kldiv``, `Kullback-Leibler divergence`_, alias=\ ``kullback_leibler``
+
    -  support multi metrics, separated by ``,``
 
 -  ``metric_freq``, default=\ ``1``, type=int
@@ -749,3 +770,9 @@ You can specific query/group id in data file now. Please refer to parameter ``gr
 .. _AUC: https://en.wikipedia.org/wiki/Receiver_operating_characteristic#Area_under_the_curve
 
 .. _log loss: https://www.kaggle.com/wiki/LogLoss
+
+.. _softmax: https://en.wikipedia.org/wiki/Softmax_function
+
+.. _One-vs-All: https://en.wikipedia.org/wiki/Multiclass_classification#One-vs.-rest
+
+.. _Kullback-Leibler divergence: https://en.wikipedia.org/wiki/Kullback%E2%80%93Leibler_divergence

include/LightGBM/config.h

@@ -363,7 +363,7 @@ struct ParameterAlias {
     {
       { "config", "config_file" },
       { "nthread", "num_threads" },
-	  { "num_thread", "num_threads" },
+      { "num_thread", "num_threads" },
       { "random_seed", "seed" },
       { "boosting", "boosting_type" },
       { "boost", "boosting_type" },
@@ -402,7 +402,7 @@ struct ParameterAlias {
       { "num_round", "num_iterations" },
       { "num_trees", "num_iterations" },
       { "num_rounds", "num_iterations" },
-	  { "num_boost_round", "num_iterations" },
+      { "num_boost_round", "num_iterations" },
       { "sub_row", "bagging_fraction" },
       { "subsample", "bagging_fraction" },
       { "subsample_freq", "bagging_freq" },
@@ -432,7 +432,7 @@ struct ParameterAlias {
       { "predict_raw_score", "is_predict_raw_score" },
       { "raw_score", "is_predict_raw_score" },
       { "leaf_index", "is_predict_leaf_index" },
-	  { "predict_leaf_index", "is_predict_leaf_index" },
+      { "predict_leaf_index", "is_predict_leaf_index" },
       { "contrib", "is_predict_contrib" },
       { "predict_contrib", "is_predict_contrib" },
       { "min_split_gain", "min_gain_to_split" },
@@ -444,7 +444,7 @@ struct ParameterAlias {
       { "bagging_fraction_seed", "bagging_seed" },
       { "workers", "machines" },
       { "nodes", "machines" },
-	  { "subsample_for_bin", "bin_construct_sample_cnt" },
+      { "subsample_for_bin", "bin_construct_sample_cnt" },
     });
     const std::unordered_set<std::string> parameter_set({
       "config", "config_file", "task", "device",

python-package/lightgbm/sklearn.py

@@ -163,7 +163,7 @@ def __init__(self, boosting_type="gbdt", num_leaves=31, max_depth=-1,
         objective : string, callable or None, optional (default=None)
             Specify the learning task and the corresponding learning objective or
             a custom objective function to be used (see note below).
-            default: 'binary' for LGBMClassifier, 'lambdarank' for LGBMRanker.
+            default: 'regression' for LGBMRegressor, 'binary' or 'multiclass' for LGBMClassifier, 'lambdarank' for LGBMRanker.
         min_split_gain : float, optional (default=0.)
             Minimum loss reduction required to make a further partition on a leaf node of the tree.
         min_child_weight : float, optional (default=1e-3)
@@ -264,7 +264,7 @@ def __init__(self, boosting_type="gbdt", num_leaves=31, max_depth=-1,
         self._best_score = None
         self._best_iteration = None
         self._other_params = {}
-        self._objective = None
+        self._objective = objective
         self._n_features = None
         self._classes = None
         self._n_classes = None
@@ -285,6 +285,8 @@ def get_params(self, deep=True):
     def set_params(self, **params):
         for key, value in params.items():
             setattr(self, key, value)
+            if hasattr(self, '_' + key):
+                setattr(self, '_' + key, value)
             self._other_params[key] = value
         return self
 
@@ -370,8 +372,6 @@ def fit(self, X, y,
         For multi-class task, the y_pred is group by class_id first, then group by row_id.
         If you want to get i-th row y_pred in j-th class, the access way is y_pred[j * num_data + i].
         """
-        if not hasattr(self, '_objective'):
-            self._objective = self.objective
         if self._objective is None:
             if isinstance(self, LGBMRegressor):
                 self._objective = "regression"
@@ -633,7 +633,8 @@ def fit(self, X, y,
         self._n_classes = len(self._classes)
         if self._n_classes > 2:
             # Switch to using a multiclass objective in the underlying LGBM instance
-            self._objective = "multiclass"
+            if self._objective != "multiclassova" and not callable(self._objective):
+                self._objective = "multiclass"
             if eval_metric == 'logloss' or eval_metric == 'binary_logloss':
                 eval_metric = "multi_logloss"
             elif eval_metric == 'error' or eval_metric == 'binary_error':

src/metric/metric.cpp

@@ -39,7 +39,7 @@ Metric* Metric::CreateMetric(const std::string& type, const MetricConfig& config
     return new MultiErrorMetric(config);
   } else if (type == std::string("xentropy") || type == std::string("cross_entropy")) {
     return new CrossEntropyMetric(config);
-  } else if (type == std::string("xentlambda")) {
+  } else if (type == std::string("xentlambda") || type == std::string("cross_entropy_lambda")) {
     return new CrossEntropyLambdaMetric(config);
   } else if (type == std::string("kldiv") || type == std::string("kullback_leibler")) {
     return new KullbackLeiblerDivergence(config);