Abstract Although the applications of data science and machine learning in solar engineering have increased tremendously in the past decade, most of the solar datasets come from heterogeneous and autonomous sources. For that reason, identifying spatially collocated and temporally aligned datasets has always been time consuming, or even frustrating sometimes, in data-driven solar research. In this regard, I present a new R package— SolarData —for easy access of some publicly available solar datasets. In version 1.0, a total of 5 datasets are included: (1) NREL physical solar model version 3; (2) NREL Oahu solar measurement grid; (3) NOAA surface radiation network; (4) SoDa Linke turbidity factor; and (5) NASA shuttle radar topography mission. This paper provides an overview of each of dataset, and gives code segments that exemplify the usage of the package. Furthermore, in the appendices, a series of self-contained R scripts are used to provide perspectives on how these datasets can be used in solar research. To promote the widespread uptake of this R package, and to facilitate future contribution and collaboration, all contents herein described are made available on GitHub. This paper is the first Data Article, a new submission type offered by the Solar Energy journal.

中文翻译：

---

SolarData：一个 R 包，用于轻松访问公开可用的太阳能数据集

摘要 尽管数据科学和机器学习在过去十年中在太阳能工程中的应用大幅增加，但大部分太阳能数据集来自异构和自主来源。出于这个原因，在数据驱动的太阳能研究中，识别空间并置和时间对齐的数据集一直很耗时，有时甚至令人沮丧。在这方面，我提出了一个新的 R 包——SolarData——用于轻松访问一些公开可用的太阳能数据集。在1.0版本中，共包含5个数据集：（1）NREL物理太阳模型版本3；(2) NREL瓦胡岛太阳能测量网格；(3) NOAA 地表辐射网络；(4) 苏打林克浊度系数；(5) NASA 航天飞机雷达地形任务。本文概述了每个数据集，并给出代码段来举例说明包的用法。此外，在附录中，一系列自包含的 R 脚本用于提供有关如何将这些数据集用于太阳能研究的观点。为了促进这个 R 包的广泛采用，并促进未来的贡献和协作，此处描述的所有内容都可以在 GitHub 上获得。这篇论文是第一篇数据文章，这是太阳能期刊提供的一种新的投稿类型。