LGBM函数及参数详解

LGBM Python API

Dataset

class lightgbm.Dataset(data, label=None, max_bin=None, reference=None, weight=None, group=None, init_score=None, silent=False, feature_name='auto', categorical_feature='auto', params=None, free_raw_data=True)

创建一个带label的训练集和交叉验证集

    trn_data = lgb.Dataset(X_tr, label=y_tr)
    val_data = lgb.Dataset(X_valid, label=y_valid)

Booster 

class lightgbm.Booster(params=None, train_set=None, model_file=None, silent=False)    #booster可用lgb来代替

• params   字典形式的参数
• train_set Training dataset.
• model_file model文件的路径
• silent 构建模型时是否打印信息

只要有lgb实例，就可以调用下列函数 ，包括 predict，

lgb.add_valid  添加交叉训练集 

lgb.attr(key)  Get attribute string from the Booster.

lgb.current_iteration()  Get the index of the current iteration.

lgb.dump_model(num_iteration=-1)   Dump Booster to jso