Contamination Map Methodology

Time series / Dataset

Traffic (PeMS San Francisco Bay, 862 sensors)

Hourly road occupancy rates (0-1) from 862 sensors on San Francisco Bay area freeways, 2015-2016 (17,544 hourly steps), from Caltrans PeMS; distributed by LSTNet and used as the long-horizon 'Traffic' dataset.

Also known as
traffic_hourly, Traffic Hourly, San Francisco Traffic, traffic.csv
Domain
transport
Homepage
pems.dot.ca.gov

Exposure by model

BenchmarkModelLevelReason
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) Chronos-2-Synth Not seen absent from the complete list of training data
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) Chronos-2 Related data seen related dataset LargeST (PeMS California) was used
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) Chronos-Bolt Unknown training data not disclosed in enough detail
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) Chronos (T5) Not seen the authors state the dataset was excluded from training
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) DADA (General Time Series Anomaly Detector with Adaptive Bottlenecks and Dual Adversarial Decoders) Unknown training data not disclosed in enough detail
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) Granite-TimeSeries-TTM-R1 Not seen the authors state the dataset was excluded from training
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) Granite-TimeSeries-TTM-R2.1 Not seen absent from the complete list of training data
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) Granite-TimeSeries-TTM-R2 Not seen absent from the complete list of training data
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) Granite-TTM-R3 Seen, split unclear the dataset was used; the source does not say which part (via GIFT-Eval Pretrain (GiftEvalPretrain))
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) Lag-Llama Seen, split unclear a training split was used; its overlap with this benchmark's test data is not established
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) Mantis (Mantis-8M) Unknown training data not disclosed in enough detail
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) Mantis+ (MantisPlus) Not seen absent from the complete list of training data
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) MantisV2 Not seen absent from the complete list of training data
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) Moirai-1.0-R Seen, split unclear a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) Moirai-1.1-R Unknown training data not disclosed in enough detail
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) Moirai-2.0-R Seen, split unclear the dataset was used; the source does not say which part (via GIFT-Eval Pretrain (GiftEvalPretrain))
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) Moirai-MoE-1.0-R Seen, split unclear a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) MOMENT-1 Training data seen only the training split was used, not the benchmark's test data (via Time Series Pile); the test portion was explicitly excluded
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) NuTime Not seen absent from the complete list of training data
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) Sundial (Timer 3.0) Unknown training data not disclosed in enough detail
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) Time-MoE Seen, split unclear the dataset was used; the source does not say which part (via Time-300B)
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) Time-RCD (TimeRCD) Not seen absent from the complete list of training data
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) Timer-S1 Seen, split unclear a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) Timer-XL (timer-base-84m) Seen, split unclear a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) Timer Seen, split unclear a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive)); this contradicts a statement that the data was excluded
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) TimeRadar Unknown training data not disclosed in enough detail
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) TimesFM 1.0 Seen, split unclear the dataset was used; the source does not say which part
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) TimesFM 2.0 Seen, split unclear the dataset was used; the source does not say which part
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) TimesFM 2.5 Seen, split unclear the dataset was used; the source does not say which part (via GIFT-Eval Pretrain (GiftEvalPretrain))
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) TimesFM 3.0 Seen, split unclear the dataset was used; the source does not say which part (via GIFT-Eval Pretrain (GiftEvalPretrain))
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) TiRex 1.1 (gifteval) Related data seen related dataset LargeST (PeMS California) was used
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) TiRex-2-fevbench (TiRex-2-f) Seen, split unclear the dataset was used; the source does not say which part
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) TiRex-2-gifteval-pretrain (TiRex-2-gp) Seen, split unclear the dataset was used; the source does not say which part (via GIFT-Eval Pretrain (GiftEvalPretrain))
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) TiRex-2-gifteval-zs (TiRex-2-g) Seen, split unclear the dataset was used; the source does not say which part
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) TiRex-2 Unknown training data not disclosed in enough detail
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) TiRex Related data seen related dataset LargeST (PeMS California) was used
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) Toto-Open-Base-1.0 Seen, split unclear the dataset was used; the source does not say which part (via GIFT-Eval Pretrain (GiftEvalPretrain))
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) Toto-2.0-2.5B-FT Seen, split unclear the dataset was used; the source does not say which part (via GIFT-Eval Pretrain (GiftEvalPretrain))
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) Toto 2.0 Not seen the authors state the dataset was excluded from training
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) Toto (2024 technical report) Seen, split unclear a training split was used; its overlap with this benchmark's test data is not established (via LOTSA (Large-scale Open Time Series Archive))
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) TSPulse r1 Not seen absent from the complete list of training data
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) TTM-R3 Seen, split unclear the dataset was used; the source does not say which part (via GIFT-Eval Pretrain (GiftEvalPretrain))
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) Tiny Time Mixer (TTM) Research-Use R2 Not seen the authors state the dataset was excluded from training
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) UniTS few-shot anomaly detection checkpoints Training data seen via base model UniTS: only the training split was used, not the benchmark's test data
Long-horizon forecasting (ETT, Electricity, Traffic, Weather, Exchange, ILI) UniTS Training data seen only the training split was used, not the benchmark's test data

Notes

Original source is Caltrans PeMS (homepage); the processed 862-series file comes from the LSTNet data repo. Aliases 'traffic_hourly' / 'Traffic Hourly' / 'San Francisco Traffic' refer to the Monash version, described as 'the San Francisco Traffic dataset used by Lai et al. (2017)' with 862 hourly series 2015-2016 (length 17,544 on forecastingdata.org, same as DLinear Table 1). The bare name 'Traffic' is ambiguous and not an alias. Not to be confused with the PEMS03/04/07/08, PEMS-Bay or LargeST datasets, which are other PeMS extracts.

Sources