removed num_class from regressor

src/estimators.jl

@@ -90,7 +90,6 @@ mutable struct LGBMRegression <: LGBMEstimator
     predict_disable_shape_check::Bool
 
     # Objective parameters
-    num_class::Int 
     is_unbalance::Bool
     boost_from_average::Bool
     reg_sqrt::Bool
@@ -332,7 +331,7 @@ function LGBMRegression(;
         linear_tree, max_bin, max_bin_by_feature, min_data_in_bin, bin_construct_sample_cnt, data_random_seed,
         is_enable_sparse, enable_bundle, use_missing, zero_as_missing, feature_pre_filter, pre_partition, categorical_feature,
         start_iteration_predict, num_iteration_predict, predict_raw_score, predict_leaf_index, predict_contrib, predict_disable_shape_check, 
-        1, is_unbalance, boost_from_average, reg_sqrt, alpha, fair_c, poisson_max_delta_step, tweedie_variance_power, 
+        is_unbalance, boost_from_average, reg_sqrt, alpha, fair_c, poisson_max_delta_step, tweedie_variance_power, 
         metric, metric_freq, is_provide_training_metric, eval_at, 
         num_machines, local_listen_port, time_out, machine_list_filename, machines, gpu_platform_id, gpu_device_id, gpu_use_dp, num_gpu,
     )

src/predict.jl

@@ -84,7 +84,7 @@ function predict(
 
     # This works the same one way or another because when n=1, (regression) reshaping is basically no-op
     # except for adding the extra dim
-    prediction = transpose(reshape(prediction, estimator.num_class, :))
+    prediction = transpose(reshape(prediction, estimator isa LGBMRegression ? 1 : estimator.num_class, :))
 
     return prediction
 

src/utils.jl

@@ -59,7 +59,9 @@ the parameters or data of the estimator whose model was saved as `filename`.
 """
 function loadmodel!(estimator::LGBMEstimator, filename::String)
     estimator.booster = LGBM_BoosterCreateFromModelfile(filename)
-    estimator.num_class = LGBM_BoosterGetNumClasses(estimator.booster)
+    if !(estimator isa LGBMRegression)
+        estimator.num_class = LGBM_BoosterGetNumClasses(estimator.booster)
+    end
     return nothing
 end
 

src/wrapper.jl

@@ -616,7 +616,7 @@ end
 
 function LGBM_BoosterGetPredict(bst::Booster, data_idx::Integer)
     out_len = Ref{Int64}()
-    num_class = LGBM_BoosterGetNumClasses(bst)
+    num_class = bst isa LGBMRegression ? 1 : LGBM_BoosterGetNumClasses(bst)
     num_data = LGBM_BoosterGetNumPredict(bst, data_idx)
     out_results = Array{Cdouble}(undef, num_class * num_data)
     @lightgbm(:LGBM_BoosterGetPredict,