Qlib quantitative source analysis: qlib/qlib/contrib/model/GBDT py

qlib gbd (lightgbm) uses an interface very similar to sklearn, and dataset.prepare returns the training set and test set.

Look at the code of the GBDT module, it first calls _prepare_data, using the data set dataset.prepare function, get two parts of train and valid, and then perform basic data inspection and transformation to fit the model.

def _prepare_data(self, dataset: DatasetH, reweighter=None) -> List[Tuple[lgb.Dataset, str]]:"""The motivation of current version is to make validation optional- train segment is necessary;"""ds_l = []assert "train" in dataset.segmentsfor key in ["train", "valid"]:if key in dataset.segments:df = dataset.prepare(key, col_set=["feature", "label"], data_key=DataHandlerLP.DK_L)if df.empty:raise ValueError("Empty data from dataset, please check your dataset config.")x, y = df["feature"], df["label"]# Lightgbm need 1D array as its labelif y.values.ndim == 2 and y.values.shape[1] == 1:y = np.squeeze(y.values)else:raise ValueError("LightGBM doesn't support multi-label training")if reweighter is None:w = Noneelif isinstance(reweighter, Reweighter):w = reweighter.reweight(df)else:raise ValueError("Unsupported reweighter type.")ds_l.append((lgb.Dataset(x.values, label=y, weight=w), key))return ds_l