BostonBusEquity / README.md
LumenscopeAI's picture
Upload README.md with huggingface_hub
54d4c82 verified

Boston Bus Equity Dataset

This dataset contains MBTA (Massachusetts Bay Transportation Authority) bus service data for Boston Bus Equity analysis.

Dataset Subsets

1. arrival_departure

Bus arrival and departure times data (2020-2026).

Schema:

  • service_date (date): Service date
  • route_id (string): Bus route identifier
  • direction_id (int): Direction (0=outbound, 1=inbound)
  • half_trip_id (string): Half trip identifier
  • stop_id (int): Bus stop identifier
  • time_point_id (string): Time point identifier
  • time_point_order (int): Order of time point on route
  • point_type (string): Type of point
  • standard_type (string): Standard type
  • scheduled (datetime): Scheduled arrival/departure time
  • actual (datetime): Actual arrival/departure time
  • scheduled_headway (float): Scheduled headway in minutes
  • headway (float): Actual headway in minutes
  • year (int): Year derived from service_date
  • month (int): Month derived from service_date

Records: ~161 million Size: ~1.9 GB

2. ridership

Bus ridership data by trip, season, route, and stop (2016-2024).

Schema:

  • season (string): Season and year (e.g., "Fall 2023")
  • route_id (string): Bus route identifier
  • route_variant (string): Route variant
  • direction_id (int): Direction
  • trip_start_time (time): Trip start time
  • day_type_id (string): Day type ID
  • day_type_name (string): Day type name (weekday, saturday, sunday)
  • stop_name (string): Bus stop name
  • stop_id (int): Bus stop identifier
  • stop_sequence (int): Stop sequence on route
  • boardings (float): Number of boardings
  • alightings (float): Number of alightings
  • load_ (float): Passenger load
  • sample_size (int): Sample size
  • year (int): Year derived from season
  • season_name (string): Season name (Fall, Spring, etc.)

Records: ~7.9 million Size: ~50 MB

3. survey

MBTA 2024 System-Wide Passenger Survey data.

Schema:

  • aggregation_level (string): Level of aggregation
  • service_mode (string): Service mode (Bus, Rail, etc.)
  • reporting_group (string): Reporting group
  • measure_group (string): Measure group
  • measure (string): Measure name
  • category (string): Category
  • weighted_percent (float): Weighted percentage
  • ObjectId (int): Object identifier
  • year (int): Survey year (2024)

Records: ~14,000 Size: ~244 KB

4. gtfs_stops

GTFS bus stop locations and information.

Schema:

  • stop_id (int): Stop identifier
  • stop_code (string): Stop code
  • stop_name (string): Stop name
  • stop_lat (float): Latitude
  • stop_lon (float): Longitude
  • zone_id (string): Fare zone
  • municipality (string): Municipality name
  • on_street (string): Street name
  • at_street (string): Cross street
  • wheelchair_boarding (int): Wheelchair accessibility
  • vehicle_type (int): Vehicle type

Records: ~10,000 Size: ~550 KB

5. gtfs_routes

GTFS route information.

Schema:

  • route_id (string): Route identifier
  • route_short_name (string): Short route name
  • route_long_name (string): Long route name
  • route_desc (string): Route description
  • route_type (int): Route type
  • route_color (string): Route color (hex)
  • line_id (string): Line identifier

Records: ~400 Size: ~25 KB

6. census (race, poverty_rates, per_capita_income)

Boston neighborhood demographic data from Census/ACS (2015-2019).

Files:

  • census_race.parquet - Race/ethnicity by neighborhood
  • census_poverty_rates.parquet - Poverty rates by neighborhood
  • census_per_capita_income.parquet - Per capita income by neighborhood

Records: ~30 per file Size: ~5-11 KB each

Usage

import pandas as pd

# Load arrival/departure data
df = pd.read_parquet("data/arrival_departure/arrival_departure.parquet")

# Load ridership data
ridership = pd.read_parquet("data/ridership/ridership.parquet")

# Load survey data
survey = pd.read_parquet("data/survey/survey.parquet")

# Load GTFS data
stops = pd.read_parquet("data/gtfs/gtfs_stops.parquet")
routes = pd.read_parquet("data/gtfs/gtfs_routes.parquet")

# Load census data
census_race = pd.read_parquet("data/census/census_race.parquet")

Data Source

License

This dataset is derived from publicly available MBTA and City of Boston data.

Citation

If you use this dataset, please cite:

@dataset{boston_bus_equity_2025,
  title={Boston Bus Equity Dataset},
  author={Boston Bus Equity Team},
  year={2025},
  publisher={HuggingFace},
  url={https://huggingface.co/datasets/LumenscopeAI/BostonBusEquity}
}