V1 photo-stimulation preprocessed structs (subset for SG-EAP Fig. 5)

Preprocessed per-session MATLAB structs from the V1 two-photon all-optical photostimulation experiments of Russell et al., Nature Communications 15, 2024. This deposit contains the 9-session subset used as the empirical anchor for Fig. 5 of State-Gated Learning Shapes the Geometry of Cooperation and Competition in Recurrent Networks (manuscript in preparation). For each session there are two structs: new_Lanimal_date.mat (neural: per-trial dF/F, ROI metadata, contrast labels, target indices) and new_BEHAV_animal_date.mat (running, pupil, lick, trial timing). The MATLAB build script newdataset_buildnewstruct_v2.m that produced the structs from the raw recordings is also included. The full Russell et al. dataset (raw imaging, behaviour, additional sessions) is not in this deposit. Please contact the original authors for the broader dataset. See the README.md for file-by-file inventory, loading examples (Python and MATLAB), and how to reproduce the Fig. 5 analysis.

本数据集包含经预处理的单会话MATLAB结构体，源自Russell等人2024年发表于《Nature Communications》第15卷的初级视觉皮层V1区双光子全光光刺激实验。本次上传仅包含其中9个会话的子集，该子集作为待投稿手稿《State-Gated Learning Shapes the Geometry of Cooperation and Competition in Recurrent Networks》中图5的实证基准数据集。针对每个会话，均包含两个结构体文件：new_Lanimal_date.mat（存储神经数据：各试次ΔF/F（dF/F）信号、ROI（Region of Interest，感兴趣区）元数据、对比度标签、目标索引）与new_BEHAV_animal_date.mat（存储行为数据：运动状态、瞳孔信号、舔舐行为、试次计时信息）。同时附带了从原始录制数据生成上述结构体的MATLAB构建脚本newdataset_buildnewstruct_v2.m。本上传未包含Russell等人的完整数据集（含原始成像数据、行为数据及额外会话数据）。如需获取更全面的数据集，请联系原论文作者。如需了解逐文件清单、Python与MATLAB环境下的数据加载示例，以及复现图5分析的具体方法，请参阅README.md文件。