Rapid Transit and Bus Prediction Accuracy Data

This file contains the prediction accuracy for subway and bus. Prediction accuracy is determined by the number of accurate predictions vs the number of total predictions for each "bin" or timeframe. Data is not guaranteed to be complete for any line or date. There is a known gap in Orange Line data from 09/02/2022 to 09/16/2022. Name Description Data Type Example weekly Date representing one week's worth of data. For both bus and subway, the week is labeled as a Friday and represents data from the previous Friday up till the Thursday the day before. (05/23/2025 represents data from 05/16/2025 to 05/22/2025.) The date is based on "service day", so "May 1" means May 1, 3:00am ET until May 2, 2:59am ET. Date 05/23/2025 mode Either "bus" for bus predictions, or "subway" for Red, Orange, Green-[B/C/D/E], Blue, and Mattapan predictions. String bus route_id The subway route the data is for. Our bus data provider does not have this data at a per-route level. String Green-B arrival_departure For bus, whether the data is about the timing of an arrival at a bus stop, or the departure from that bus stop. Bus only supports "departure". Absent on subway data because subway uses a "blended" approach of departure predictions at terminals, and arrival predictions otherwise. String departure bin The bin a prediction belongs to based on how far in the future the predicted event is for. The options are "0-3 min", "3-6 min", "6-12 min", and "12-30 min". String 0-3 min num_predictions The count of predictions sampled that meet the criteria of the other fields. Integer 50000 num_accurate_predictions Of the num_predictions, how many of them were considered accurate, where "accurate" means the predicted number of seconds was within a threshold of the actual number of seconds, based on the bin. For a given bin, the passing threshold is if a vehicle arrives: 0-3 min: 60 seconds early to 60 seconds late, 3-6 min: 90 seconds early to 120 seconds late, 6-12 min: 150 seconds early to 210 seconds late, 12:30 min: 240 seconds early to 360 seconds late. Integer 30000 MassDOT/MBTA shall not be held liable for any errors in this data. This includes errors of omission, commission, errors concerning the content of the data, and relative and positional accuracy of the data. This data cannot be construed to be a legal document. Primary sources from which this data was compiled must be consulted for verification of information contained in this data.本文件收录地铁与公交的预测准确率数据。预测准确率的计算逻辑为：针对每个「时段区间（bin）」或时间框架，以准确预测次数与总预测次数的比值作为衡量标准。本数据不保证任意线路或日期的数据完整，已知2022年9月2日至2022年9月16日期间橙线数据存在缺失。 ### 字段：weekly **字段说明**：代表一周数据的日期标签。公交与地铁数据的周标签均设置为周五，对应的数据周期为前一个周五至当周周四（例如2025年5月23日对应2025年5月16日至2025年5月22日的数据）。该日期基于「服务日（service day）」定义，即标注为「5月1日」的时段，指代美国东部时间（ET）5月1日3:00至5月2日2:59的区间。 **数据类型**：日期（Date） **示例值**：05/23/2025 ### 字段：mode **字段说明**：数据所属的交通方式，可选值为「bus（公交）」与「subway（地铁）」。其中地铁涵盖红线、橙线、绿线（含B/C/D/E支线）、蓝线以及马特潘线（Mattapan）的预测数据。 **数据类型**：字符串（String） **示例值**：bus ### 字段：route_id **字段说明**：对应数据所属的地铁线路。公交数据供应商未提供单线路维度的该类数据。 **数据类型**：字符串（String） **示例值**：Green-B ### 字段：arrival_departure **字段说明**：仅针对公交数据，用于标识数据为公交站点的到站时刻预测还是离站时刻预测。公交场景下仅支持「departure（离站）」。地铁数据无此字段，因地铁采用「混合（blended）」预测模式：在终点站使用离站预测，其余站点使用到站预测。 **数据类型**：字符串（String） **示例值**：departure ### 字段：bin **字段说明**：根据预测事件的未来时长划分的时段区间，可选值为「0-3 min」、「3-6 min」、「6-12 min」以及「12-30 min」。 **数据类型**：字符串（String） **示例值**：0-3 min ### 字段：num_predictions **字段说明**：符合当前字段筛选条件的采样预测总次数。 **数据类型**：整数（Integer） **示例值**：50000 ### 字段：num_accurate_predictions **字段说明**：在`num_predictions`的总预测次数中，被判定为准确的预测次数。「准确」的定义为：预测的秒数与实际秒数的偏差处于对应时段区间的阈值范围内。各时段区间的合格阈值如下： - 0-3分钟：提前60秒至延迟60秒 - 3-6分钟：提前90秒至延迟120秒 - 6-12分钟：提前150秒至延迟210秒 - 12-30分钟：提前240秒至延迟360秒 **数据类型**：整数（Integer） **示例值**：30000 马萨诸塞州交通部（MassDOT）/马萨诸塞湾交通局（MBTA）不对本数据中的任何错误承担责任，包括疏漏错误、不当作为错误、数据内容错误以及数据的相对精度与位置精度错误。本数据不得被视为法律文件，如需验证本数据包含的信息，需查阅本数据所基于的原始数据源。