遇见数据集

UniverseTBD/mmu_jwst_gdn_grizli_v7.3_all_96

收藏
Hugging Face2026-05-20 更新2026-03-29 收录
官方服务:

资源简介:

这是一个名为mmu_jwst_gdn_grizli_v7.3_all_96的HATS目录集合,代表该特定天文数据集。该数据集是Multimodal Universe(多模态宇宙)的一部分,这是一个大规模的多模态天文数据集合,旨在支持大规模机器学习应用,相关详细信息可参考论文《The Multimodal Universe: Enabling Large-Scale Machine Learning with 100TBs of Astronomical Scientific Data》。数据集包含68,308行和815列,分为12个分区,占用磁盘空间10.2 GiB,由hats-import v0.7.3和hats v0.7.3构建。列包括天文观测相关的多种指标,如不同波段的流量(例如f090w、f105w等)、背景值、校正参数、标志位和掩码等,适用于天文数据分析和交叉匹配任务。数据集以Apache Parquet格式存储,可通过LSDB、pandas、pyarrow等工具访问。

This is a HATS catalog collection named mmu_jwst_gdn_grizli_v7.3_all_96, representing this specific astronomical dataset. This dataset is part of the Multimodal Universe, a large-scale multimodal astronomical data collection designed to support large-scale machine learning applications. Detailed relevant information can be found in the paper titled *The Multimodal Universe: Enabling Large-Scale Machine Learning with 100TBs of Astronomical Scientific Data*. This dataset contains 68,308 rows and 815 columns, is divided into 12 partitions, occupies 10.2 GiB of disk space, and is constructed using hats-import v0.7.3 and hats v0.7.3. Its columns include various metrics related to astronomical observations, such as fluxes in different bands (e.g., f090w, f105w, etc.), background values, correction parameters, flag bits, and masks, and is suitable for astronomical data analysis and cross-matching tasks. This dataset is stored in Apache Parquet format and can be accessed via tools such as LSDB, pandas, pyarrow, and others.

提供机构:
UniverseTBD
二维码
社区交流群
二维码
科研交流群
商业服务