UniverseTBD/mmu_jwst_primer_cosmos

该数据集是mmu_jwst_primer_cosmos HATS目录集合，属于多模态宇宙项目的一部分，是一个大规模的多模态天文数据集合。数据集基于詹姆斯·韦伯太空望远镜（JWST）的近红外相机观测数据，来自早期深场巡天。数据模态包括以光度测量目录中的源为中心的固定尺寸图像切块（96×96像素），图像为多波段，覆盖6或7个滤波器，波长范围约为0.9μm至4.4μm。这些JWST深场巡天图像已用于大量科学出版物，包括机器学习应用。注意事项：不同巡天具有不同的波长覆盖范围，缺失的波段用零数组表示以简化数据加载。数据属于公共领域，使用了从黎明JWST档案（DJA）检索的数据产品，该档案是宇宙黎明中心（DAWN）的一项倡议。数据集包含51,058行和11列，分区数为15，磁盘大小为15.1 GiB，采用HATS目录格式，可通过LSDB Python框架访问和跨目录匹配。

This dataset is the mmu_jwst_primer_cosmos HATS Catalog Collection, part of the Multimodal Universe project, a large-scale collection of multimodal astronomical data. It is based on James Webb Space Telescope (JWST) NIRCam observations from early deep field surveys. The data modality consists of fixed-size image cutouts (96×96 pixels) centered on sources from photometric catalogs, with multi-band images covering 6 or 7 filters across wavelengths from approximately 0.9μm to 4.4μm. Images from these JWST deep field surveys have been used in numerous scientific publications, including machine learning applications. Caveats: Different surveys have different wavelength coverage, and missing bands are represented as arrays of zeros to simplify data loading. The data are in the public domain, using data products retrieved from the Dawn JWST Archive (DJA), an initiative of the Cosmic Dawn Center (DAWN). The dataset contains 51,058 rows and 11 columns, with 15 partitions, a disk size of 15.1 GiB, in HATS catalog format, accessible via the LSDB Python framework for cross-matching.