Pedal Poll / Sondo Vélo 2021-2024 analysis

This repo contains the code and documentation for analysis of the 2022-2024 Pedal Poll / Sondo Vélo data. If you're looking for the code to generate infographics and interactive tables, go to this repo instead!

The 2021-2024 analysis paper can be found here.

The 2021 results can be found here and the associated academic paper is here.

Cleaning and analysis steps:

Data

This repository contains all raw data from Pedal Poll used in the analysis. The following external datasets are also required to reproduce the analysis:

1. Pedal Poll 2021 cleaned data. In all scripts this data is loaded directly from the online version, which can be found here.
2. Spatial Access Measures: a database of accessibility to destinations produced by Statistics Canada. The file can be found here.
3. 2021 census dissemination block boundary file: this is used to assign count locations to dissemination blocks to match with Spatial Access Measures data. The file can be found here.

Index of files:

• 20XX_data_dictionary.csv: data dictionary for cleaned data files produced from data_cleaning_2022.ipynb and data_cleaning.ipynb. Because the 2022 data is formatted like the 2021 data and has a different format from 2023 and 2024, there are two data dictionaries.
• 20XX_hand_count.csv: paper count data from 2022, 2023, and 2024.
• 20240916_153200.csv: raw CounterPoint app data from 2023, downloaded with this query.
• 20250624_142657.csv: raw CounterPoint app data from 2024, downloaded with this query.
• CounterPoint Data Explorer - Bicycle Demographics Submission Data_Page 1_Table.csv: raw CounterPoint app data from 2022, downloaded with the now-replaced CounterPoint data explorer.
• Official Count Locations IDs.csv: list of official 2021 focus community count sites, used only to preferentially label grouped CounterPoint IDs where possible.
• counterpoint_official_sites_multiyear_counts-CanBICS-v2_sheet1.csv: unique counterpoints classified after 2023 count.
• PP_CounterPoints_Cycle_Facilities_FULL_02JAN2025.csv: list of unique CounterPoints and classified Can-BICS infrastructure in each count year.
• brampton_2023_paper.csv: additional 2023 paper count data from Brampton.
• cma_population_2021.csv: uPsed to determine community population sizes for Discussion. Processed file created from 2021 census data downloaded on 2025-03-03 from https://www12.statcan.gc.ca/census-recensement/2021/dp-pd/prof/index.cfm?Lang=E
• hand_count_2022_unique_counterpoints_filled.csv: 2022 paper count data with CounterPoint IDs manually filled in where possible.

Scripts

The following scripts run in this order reproduce the analysis and figures for the paper.

1. data_cleaning_2022.ipynb and data_cleaning.ipynb: scripts to clean data for 2022 and 2023/2024.
2. counterpoint_location_grouping.ipynb: group CounterPoint pins within 100 m to better compare counts over multiple years.
3. Can-BICS_classification.ipynb: sort CounterPoint locations for classification by hand with Can-BICS.
4. analysis_processing.ipynb: load all years of cleaned datasets and group counts by count session, calculate bikes per hour.
5. regression_model.ipynb: add Spatial Access Measures data and process data for regression model.
6. regression.R: run regression model and calculate confidence intervals.
7. paper_figures.ipynb: produce figures and tables for paper.