Generalized Independence Noise (GIN) condition-based method

Algorithm Introduction

Learning the structure of Linear, Non-Gaussian LAtent variable Model (LiNLAM) based the GIN 1 condition.

Usage

from causal-learn.search.FCMBased.GIN.GIN import GIN
G, K = GIN(data)

Parameters

data: numpy ndarray. Data set.

Returns

G: GeneralGraph. Causal graph.

K: list. Causal Order.

1

Xie, F., Cai, R., Huang, B., Glymour, C., Hao, Z., & Zhang, K. (2020, January). Generalized Independent Noise Condition for Estimating Latent Variable Causal Graphs. In NeurIPS.