CSE472-blanket-challenge/phase1-dataset

Phase 1 Dataset是一个包含随机生成的有向无环图（DAG）和观测数据的因果发现数据集。每个数据集包含一个独特的图标识符、变量数、边数、图密度、目标变量索引、邻接矩阵、马尔可夫毯成员标记和数据样本。数据集通过线性高斯结构方程模型（SCM）生成，包含200个样本。图是随机生成的，目标变量是随机选择的非根节点，确保具有因果依赖。

Phase 1 Dataset is a causal discovery dataset containing randomly generated directed acyclic graphs (DAGs) and observational data. Each dataset includes a unique graph identifier, number of variables, number of edges, graph density, index of the target variable, adjacency matrix, Markov blanket member mask, and data samples. The datasets are generated through a linear Gaussian structural equation model (SCM) with 200 samples. The graphs are randomly generated, and the target variable is randomly selected as a non-root node, ensuring causal dependencies.