Dataset

This dataset was created by databender

Contents

This dataset was created to analyze the growth and development of metro systems across Indian cities and various cities around the globe. By gathering data from various online sources, including Wikipedia, I aimed to explore patterns in metro network expansion, ridership trends, and system characteristics. The process involved extensive data cleaning to remove inconsistencies and ensure clarity, making it ready for exploratory data analysis (EDA) and visualization. The goal was to gain valuable insights into metro systems' performance, growth trajectory, and factors influencing their success, providing a foundation for future urban transportation planning and analysis.

This line layer was created from the GTFS data feeds from the Washington D.C. Metropolitan Area Transportation Authority (WMATA) to represent the new WMATA bus transit routes for the Better Bus Network Redesign(BBNR). Lines in this layer represent individual bus routes; they were generalized from the GTFS format where lines depicted individual services. On June 29, 2025, WMATA implemented a brand new Metrobus Network – known as Better Bus. Read more about this update from BBNR info page. Each line represents the route that a specific bus follows during its daily service. The unique field, shape_id, is derived from the GTFS to identify the route, while route_id (route_shortname), direction, and variation fields use familiar letter or number designations for buses, with distinct IDs for each route. This layer was developed as part of the WMATA Better Bus Network Redesign and will be updated as necessary to reflect changes in the transit system. Learn more about GTFS feeds at wmata.com/about/developers.The District Department of Transportation (DDOT) has added DC attribution for Roadway Blocks and Roadway Subblocks to the Metro Bus Stop dataset. Use BLOCKKEY, SUBBLOCKKEY, and ROUTEID to relate back to DC government data. Learn more about DDOT's roadway centerline data at opendata.dc.gov/pages/roadway-centerlines.

The dataset contains lines representing Metro lines in the Washington DC Metropolitan area. Lines were taken from legacy data from WMATA and fit to orthophotography and extracted planimetric data.

FBI National Incident-Based Reporting System (FBI NIBRS) crime data for Metro Transit Police (Other) in District of Columbia, including incidents, statistics, demographics, and detailed incident information.

Average daily Metrobus ridership using a quarter-mile grid. A quarter mile is often used as the walkable distance to and from a bus stop based on a 15-minute walk. The highest ridership areas that are shown in red grid cells are indicative of the jobs, population and activity generators within and near those grid cells.

Context

The [Metro Rail Transit Line 3](https://en.wikipedia.org/wiki/MRT_Line_3_(Metro_Manila), also known as the MRT Line 3, MRT-3, or Metrostar Express, is a rapid transit line in Metro Manila in the Philippines. The line runs in an orbital north to south route following the alignment of Epifanio de los Santos Avenue (EDSA). Despite its name, the line is more akin to a light rapid transit system owing to its tram-like rolling stock while having total grade separation and high passenger throughput. The line is officially known as the Yellow Line.

Content

The dataset is taken from my FOI request. If you want the latest data, you have to request again from them.

I've also provided a cleaned and tidy dataset in wide format from the following script: ``` import pandas as pd import os import re # For extracting the year from sheet names

Define file paths

excel_path = # specify the xlsx location output_path = # specify the output path

Read all sheets into a dictionary of DataFrames

sheets = pd.read_excel(excel_path, sheet_name=None)

Inspect sheet names

print("Sheets found:", sheets.keys())

df_list = [] for sheet_name, df in sheets.items(): # Extract year from sheet name (expects "DAILY YYYY" format) match = re.search(r"\d{4}", sheet_name) # Finds a 4-digit year year = int(match.group()) if match else None # Convert to integer

if year is None:
  print(f"Warning: Could not extract year from sheet name '{sheet_name}'")
  continue # Skip this sheet if year is missing

df["Year"] = year # Add a column to store the extracted year

# Print first few rows for debugging
print(f"

First few rows of {sheet_name} (Year {year}):") print(df.head())

df_list.append(df)

df = pd.concat(df_list, ignore_index=True)

Drop "Unnamed" columns (empty column headers)

df = df.loc[:, ~df.columns.str.contains("Unnamed")]

Ensure first row isn't mistakenly treated as data instead of column names

if df.columns[0] != "Date": df.columns = df.iloc[0] # Set first row as column names df = df[1:] # Drop the now duplicated first row

Convert ridership columns to numeric (handling errors like #DIV/0!)

for col in df.columns: if col not in ["Date", "Year"]: # Skip non-ridership columns df[col] = pd.to_numeric(df[col], errors="coerce")

Check for any remaining non-numeric columns

print(" Non-numeric columns:", df.select_dtypes(exclude=["number"]).columns.tolist())

Remove rows where the railway was non-operational (i.e., all ridership values are NaN)

df = df.dropna(how="all", subset=df.columns[1:]) # Ignore 'Date' column in filtering

Fill missing values with 0

df.fillna(0, inplace=True)

Save cleaned dataset

os.makedirs(os.path.dirname(output_path), exist_ok=True) # Ensure directory exists df.to_csv(output_path, index=False)

print(f" Cleaned dataset saved to: {output_path}") print("Final dataset preview:") print(df.head()) ``` Warning: You have to remove unrelated texts such as the title and values outside the table for the script to run.

Acknowledgements 1. Department of Transportation (DOTr) - Philippines 2. Metro Rail Transit Corporation

The data provided in this dataset is preliminary in nature and may have not been investigated by a detective at the time of download. The data is therefore subject to change after a complete investigation. This data represents only calls for police service where a police incident report was taken. Due to the variations in local laws and ordinances involving crimes across the nation, whether another agency utilizes Uniform Crime Report (UCR) or National Incident Based Reporting System (NIBRS) guidelines, and the results learned after an official investigation, comparisons should not be made between the statistics generated with this dataset to any other official police reports. Totals in the database may vary considerably from official totals following the investigation and final categorization of a crime. Therefore, the data should not be used for comparisons with Uniform Crime Report or other summary statistics.Data is broken out by year into separate CSV files. Note the file grouping by year is based on the crime's Date Reported (not the Date Occurred).Older cases found in the 2003 data are indicative of cold case research. Older cases are entered into the Police database system and tracked but dates and times of the original case are maintained.Data may also be viewed off-site in map form for just the last 6 months on communitycrimemap.comData Dictionary:

Field Name

Field Description

Incident Number

the number associated with either the incident or used as reference to store the items in our evidence rooms

Date Reported

the date the incident was reported to LMPD

Date Occurred

the date the incident actually occurred

Badge ID

Badge ID of responding Officer

Offense Classification

NIBRS Reporting category for the criminal act committed

Offense Code Name

NIBRS Reporting code for the criminal act committed

NIBRS_CODE

the code that follows the guidelines of the National Incident Based Reporting System. For more details visit https://ucr.fbi.gov/nibrs/2011/resources/nibrs-offense-codes/view

NIBRS Group

hierarchy that follows the guidelines of the FBI National Incident Based Reporting System

Was Offense Completed

Status indicating whether the incident was an attempted crime or a completed crime.

LMPD Division

the LMPD division in which the incident actually occurred

LMPD Beat

the LMPD beat in which the incident actually occurred

Location Category

the type of location in which the incident occurred (e.g. Restaurant)

Block Address

the location the incident occurred

City

the city associated to the incident block location

Zip Code

the zip code associated to the incident block location

Contact:LMPD Open Records lmpdopenrecords@louisvilleky.gov

Metro Performance and Accountability Data

In financial year 2023, more than *** million passenger journeys took place in metro in Delhi daily. They were a huge increase in comparison with the previous year. The huge decrease in 2021 was due to the impact of the Covid-19 pandemic. Delhi was one of the most congested cities in the world. Several new metro routes were under construction, and the government was planning to invest more in the public transport section, which would increase passenger traffic and improve the traffic congestion.

Maryland Transit - Metro Subway StationsMaryland Transit Administration Metro Subway Stations. Ridership data is based MTA"s Fiscal Year 2024. Data last updated: 11/2024. See https://mta.maryland.gov/metro-subway for more information.This is a MD iMAP hosted service layer. Find more information at https://imap.maryland.gov.Feature Service Layer Link:https://mdgeodata.md.gov/imap/rest/services/Transportation/MD_Transit/FeatureServer/4

This is daily passenger statistics for foreigners using single-ride tickets on Seoul Metro Lines 1-8. Data is uploaded annually and consists of serial number, date of transport, line, station name, boarding and alighting category, ticket type, passenger type, and number of passengers by time zone. (Data from January 2025 to June 2025.) Station name: Seoul Metro Lines 1-8 stations Boarding and alighting category: Boarding, alighting Ticket type: Single-use transportation card Passenger type: English general, English child, Japanese general, Japanese child, Chinese general, Chinese child Time zone example: 04:00 > 04:00:00~04:59:59

FBI National Incident-Based Reporting System (FBI NIBRS) crime data for Metropolitan Transit Commission (Other) in Minnesota, including incidents, statistics, demographics, and detailed incident information.

***NOTE: There are problems with data accuracy on bus ridership counts for June 2019, July 2019, and August 2019 as Phoenix and Valley Metro attempt to integrate new hardware with the existing system. For Mesa this is clearly impacting routes 156, 531, and 541, however, there are minor anomalies with some of the other routes.

Breakdown of bus service ridership on a monthly basis per route in Mesa (data lags for approximately 30 days) and data is provided by Valley Metro, an external agency who is responsible for both bus service and light rail in Mesa. For a regional view of both bus service and light rail ridership, please go to their website, under their Publications/Reports section. Valley Metro tracks bus riders as they board. Bus service expands the scope and range of light rail, in addition to being a geographically more flexible means of public transportation in comparison to light rail.

The data set captures information starting in July 2013. Effective October 24, 2016, AZ Link has been consolidated with route 112, and the Main Street Link consolidated with route 40.

Data reporting can be delayed up to 60 days.

As of 2025, the Delhi National Capital Region had the largest operational metro network in India with a rail length of *** kilometers. It was followed by Mumbai with **** kilometers and Bengaluru with ** kilometers. India’s urban rail transportation consisted of suburban rail, which was operated by Indian Railways, and rapid transit, also known as metro, mostly operated by the respective local metro corporations.

The Bangalore Metro Rail Corporation Limited (BMRCL) publishes daily ridership data every 24 hours. Unfortunately, they do not provide historical data beyond one day. I have been collecting ridership data from the BMRCL website since 26th October, 2024 and will preserve it here for anyone who wishes to analyse this data. As the dataset evolves over time, it will lend itself to deeper analysis of metro traffic, ridership and access patterns.

Metro Lines (regional). The dataset contains locations and attributes of Metro lines, created as part of the DC Geographic Information System (DC GIS) for the D.C. Office of the Chief Technology Officer (OCTO) by participating D.C. government agencies. Lines were taken from legacy data from WMATA and fit to orthophotography and extracted planimetric data. This data was downloaded by DoIT staff from http://dcatlas.dcgis.dc.gov/catalog/download.asp?downloadID=2170&downloadTYPE=ESRI.This is a MD iMAP hosted service layer. Find more information at https://imap.maryland.gov.Feature Service Layer Link:https://mdgeodata.md.gov/imap/rest/services/Transportation/MD_Transit/FeatureServer/8

This data view shows the proximity to public transportation, and modal share of commuters by metropolitan city.

This dataset includes all 7 metro counties that have made their parcel data freely available without a license or fees.

This dataset is a compilation of tax parcel polygon and point layers assembled into a common coordinate system from Twin Cities, Minnesota metropolitan area counties. No attempt has been made to edgematch or rubbersheet between counties. A standard set of attribute fields is included for each county. The attributes are the same for the polygon and points layers. Not all attributes are populated for all counties.

NOTICE: The standard set of attributes changed to the MN Parcel Data Transfer Standard on 1/1/2019.
https://www.mngeo.state.mn.us/committee/standards/parcel_attrib/parcel_attrib.html

See section 5 of the metadata for an attribute summary.

Detailed information about the attributes can be found in the Metro Regional Parcel Attributes document.

The polygon layer contains one record for each real estate/tax parcel polygon within each county's parcel dataset. Some counties have polygons for each individual condominium, and others do not. (See Completeness in Section 2 of the metadata for more information.) The points layer includes the same attribute fields as the polygon dataset. The points are intended to provide information in situations where multiple tax parcels are represented by a single polygon. One primary example of this is the condominium, though some counties stacked polygons for condos. Condominiums, by definition, are legally owned as individual, taxed real estate units. Records for condominiums may not show up in the polygon dataset. The points for the point dataset often will be randomly placed or stacked within the parcel polygon with which they are associated.

The polygon layer is broken into individual county shape files. The points layer is provided as both individual county files and as one file for the entire metro area.

In many places a one-to-one relationship does not exist between these parcel polygons or points and the actual buildings or occupancy units that lie within them. There may be many buildings on one parcel and there may be many occupancy units (e.g. apartments, stores or offices) within each building. Additionally, no information exists within this dataset about residents of parcels. Parcel owner and taxpayer information exists for many, but not all counties.

This is a MetroGIS Regionally Endorsed dataset.

Additional information may be available from each county at the links listed below. Also, any questions or comments about suspected errors or omissions in this dataset can be addressed to the contact person at each individual county.

Anoka = http://www.anokacounty.us/315/GIS
Caver = http://www.co.carver.mn.us/GIS
Dakota = http://www.co.dakota.mn.us/homeproperty/propertymaps/pages/default.aspx
Hennepin = https://gis-hennepin.hub.arcgis.com/pages/open-data
Ramsey = https://www.ramseycounty.us/your-government/open-government/research-data
Scott = http://opendata.gis.co.scott.mn.us/
Washington: http://www.co.washington.mn.us/index.aspx?NID=1606