[docs] document CLI behavior when label_column is omitted (#4485)

docs/Parameters.rst

@@ -762,6 +762,8 @@ Dataset Parameters
 
    -  add a prefix ``name:`` for column name, e.g. ``label=name:is_click``
 
+   -  if omitted, the first column in the training data is used as the label
+
    -  **Note**: works only in case of loading data directly from file
 
 -  ``weight_column`` :raw-html:`<a id="weight_column" title="Permalink to this parameter" href="#weight_column">&#x1F517;&#xFE0E;</a>`, default = ``""``, type = int or string, aliases: ``weight``

examples/binary_classification/train.conf

@@ -26,6 +26,9 @@ metric_freq = 1
 # true if need output metric for training data, alias: tranining_metric, train_metric
 is_training_metric = true
 
+# column in data to use as label
+label_column = 0
+
 # number of bins for feature bucket, 255 is a recommend setting, it can save memories, and also has good accuracy. 
 max_bin = 255
 

examples/lambdarank/train.conf

@@ -29,6 +29,9 @@ metric_freq = 1
 # true if need output metric for training data, alias: tranining_metric, train_metric
 is_training_metric = true
 
+# column in data to use as label
+label_column = 0
+
 # number of bins for feature bucket, 255 is a recommend setting, it can save memories, and also has good accuracy. 
 max_bin = 255
 

examples/multiclass_classification/train.conf

@@ -41,6 +41,9 @@ metric_freq = 1
 # true if need output metric for training data, alias: tranining_metric, train_metric
 is_training_metric = true
 
+# column in data to use as label
+label_column = 0
+
 # number of bins for feature bucket, 255 is a recommend setting, it can save memories, and also has good accuracy. 
 max_bin = 255
 

examples/parallel_learning/train.conf

@@ -26,6 +26,9 @@ metric_freq = 1
 # true if need output metric for training data, alias: tranining_metric, train_metric
 is_training_metric = true
 
+# column in data to use as label
+label_column = 0
+
 # number of bins for feature bucket, 255 is a recommend setting, it can save memories, and also has good accuracy. 
 max_bin = 255
 

examples/regression/train.conf

@@ -26,6 +26,9 @@ metric_freq = 1
 # true if need output metric for training data, alias: tranining_metric, train_metric
 is_training_metric = true
 
+# column in data to use as label
+label_column = 0
+
 # number of bins for feature bucket, 255 is a recommend setting, it can save memories, and also has good accuracy. 
 max_bin = 255
 

examples/xendcg/train.conf

@@ -29,6 +29,9 @@ metric_freq = 1
 # true if need output metric for training data, alias: tranining_metric, train_metric
 is_training_metric = true
 
+# column in data to use as label
+label_column = 0
+
 # number of bins for feature bucket, 255 is a recommend setting, it can save memories, and also has good accuracy. 
 max_bin = 255
 

include/LightGBM/config.h

@@ -657,6 +657,7 @@ struct Config {
   // desc = used to specify the label column
   // desc = use number for index, e.g. ``label=0`` means column\_0 is the label
   // desc = add a prefix ``name:`` for column name, e.g. ``label=name:is_click``
+  // desc = if omitted, the first column in the training data is used as the label
   // desc = **Note**: works only in case of loading data directly from file
   std::string label_column = "";