Data. A machine learning approach to fill gaps in dendrometer data

The dataset consists of hourly dendrometer-derived tree growth data prepared for gap filling with Extreme Gradient Boosting (XGBoost). Each record includes the temporal predictors day of year (DOY), year, and hour of day (0–23), together with the response variable Label, which represents the GRO output from the zero-growth model. The temporal variables must be continuous and contain no missing values, while missing values in Label indicate the data gaps to be predicted.

本数据集包含由树木径向生长仪（dendrometer）采集得到的逐小时树木生长数据，该数据已针对使用极端梯度提升树（Extreme Gradient Boosting, XGBoost）进行缺口填充的任务完成预处理。每条记录均包含三类时间型预测变量：年中日（day of year, DOY）、年份以及当日小时（取值范围0–23），同时附带响应变量Label，该变量代表零生长模型（zero-growth model）输出的GRO值。时间型变量需为连续型数据且无缺失值，而Label字段的缺失值即代表待预测的数据缺口。