Calibrated block-group level coefficients for NYS mode choice

Calibrated block-group level coefficients for New York State mode choice. The coefficients are calibrated by a group-level agent-based mixed logit (GLAM logit) model using Replica's synthetic datasets. This .csv file contains 120,740 rows. Each row contains a set of mode choice coefficients for each block-group OD pair and each population segment (we call this an agent). The empirical distribution of agent-level coefficients is neither Gumbel nor Gaussian, which reveals a regional divergence of the value of time and mode preference, indicating potential inequity issues in the transportation system. This is infeasible for conventional discrete choice models (DCMs) to capture.本数据集为纽约州出行方式选择的街区组（block-group）层面校准系数。该系数通过基于群体的智能体混合Logit（GLAM Logit）模型，依托Replica合成数据集完成校准。本CSV文件共包含120,740行数据，每一行对应一组街区组OD（起点-终点，Origin-Destination）对与每一类人口细分群体（本研究称其为智能体（agent））的出行方式选择系数。智能体层面系数的经验分布既不服从Gumbel分布也不服从高斯分布，这揭示了出行时间价值与出行方式偏好的区域分化，反映出交通系统中潜在的公平性问题，而传统离散选择模型（DCMs）难以捕捉该特征。