coOpenData provides a lightweight R interface to the Colorado Open Data Portal.
The package allows users to search, filter, and download datasets from the Colorado Open Data Portal directly into R without manually constructing API queries, handling JSON responses, or performing type conversion.
Designed for students, educators, researchers, journalists, civic technologists, and analysts, coOpenData reduces the technical overhead required to begin working with municipal Open Data while preserving access to the underlying Socrata infrastructure.
How coOpenData Works
The package provides a streamlined interface to the Colorado Open Data Portal’s Socrata API.
Internally, coOpenData:
- retrieves metadata from the live Colorado Open Data catalog
- constructs parameterized HTTP requests
- downloads JSON responses from Socrata endpoints
- converts results into tidy tibble outputs
- optionally cleans column names
- optionally performs conservative type coercion
Most workflows begin with co_list_datasets(), which retrieves a live catalog of datasets available through the Colorado Open Data Portal.
Datasets can then be downloaded using either:
- a human-readable catalog
key - the official Socrata dataset U_ID, such as
"e4mf-ggwf"
The human-readable key is designed to improve readability and usability, while the U_ID is the stable identifier used by the Socrata platform.
Core Functions
The package provides three primary functions:
co_list_datasets()retrieves a live catalog of available Colorado Open Data datasets, including human-readable keys, Socrata U_IDs, names, and other available metadata.co_pull_dataset()downloads cataloged datasets using either a human-readable key or Socrata U_ID, with support for filtering, ordering, date ranges, optional column-name cleaning, and optional type coercion.co_any_dataset()downloads data directly from a valid Socrata JSON endpoint without requiring the dataset to appear in the package catalog.
Datasets retrieved through co_pull_dataset() support arguments including:
limitfiltersdatefromtodate_fieldwhereorderclean_namescoerce_types
All functions return tibble outputs.
Advanced users may also provide raw SoQL conditions through the where argument.
SoQL, or Socrata Query Language, is the query syntax used by Socrata-powered Open Data portals. Additional information is available from the Socrata developer documentation.
Installation
Install from CRAN
install.packages("coOpenData")Install the development version from GitHub
# install.packages("pak")
pak::pak("gomes-sh/coOpenData")Alternatively:
# install.packages("remotes")
remotes::install_github("gomes-sh/coOpenData")Example
library(coOpenData)
library(dplyr)
# Browse available datasets
catalog <- co_list_datasets()
# Search for datasets containing a keyword
catalog |>
filter(grepl("KEYWORD", name, ignore.case = TRUE)) |>
select(key, uid, name)
# Pull a dataset using its U_ID
example_data <- co_pull_dataset(
dataset = "e4mf-ggwf",
limit = 100
)
# Pull the same dataset using its catalog key
example_data_by_key <- co_pull_dataset(
dataset = "data.colorado.gov",
limit = 100
)
# Pull filtered data
filtered_data <- co_pull_dataset(
dataset = "e4mf-ggwf",
limit = 100,
filters = list(
grade_levels = "ECE-5"
)
)The filters argument accepts a named list and automatically constructs the corresponding SoQL filtering conditions.
Multiple values may be supplied for one field:
filtered_data <- co_pull_dataset(
dataset = "e4mf-ggwf",
limit = 100,
filters = list(
grade_levels = c("ECE-5", "K-5")
)
)Multiple fields may also be combined:
filtered_data <- co_pull_dataset(
dataset = "e4mf-ggwf",
limit = 100,
filters = list(
grade_levels = "ECE-5",
classification = "Charter"
)
)Date filtering is available for datasets containing date or datetime fields:
date_filtered_data <- co_pull_dataset(
dataset = "e4mf-ggwf",
from = "2023-01-01",
to = "2024-01-01",
date_field = "last_verified",
limit = 100
)Accessing Any Socrata Endpoint
When a dataset is not available through co_list_datasets(), it can be downloaded directly using co_any_dataset().
endpoint_data <- co_any_dataset(
json_link = "https://data.colorado.gov/resource/e4mf-ggwf.json",
limit = 100
)Use co_pull_dataset() for catalog-based workflows and co_any_dataset() when working directly with a Socrata JSON endpoint.
Learn by Example
A complete introductory workflow is available in the package vignette:
vignette("getting-started", package = "coOpenData")The vignette demonstrates how to:
- browse the dataset catalog
- download data using a key or U_ID
- filter records
- work with date ranges
- access direct JSON endpoints
- perform a simple analysis
Package Website
Complete documentation is available on the package website:
https://nyc-open-data-lab.github.io/coOpenData/
The website includes:
- function reference pages
- installation instructions
- introductory articles
- vignettes
- release notes
Development
To run the package tests locally:
devtools::test()To rebuild the documentation:
devtools::document()To run a complete package check:
devtools::check()To rebuild the pkgdown website:
pkgdown::build_site()Contributing
Contributions are welcome.
To report a bug, request a feature, or suggest an improvement, open an issue on GitHub:
https://github.com/nyc-open-data-lab/coOpenData/issues
Pull requests are also welcome. Before submitting a pull request, please ensure that:
- package documentation has been regenerated
- automated tests pass
-
devtools::check()completes successfully - new behavior is documented and tested
Maintenance
Because the package retrieves metadata dynamically from the live Colorado Open Data catalog, newly published datasets may become available without requiring a package update.
Package updates may still be required when the portal changes its catalog structure, dataset metadata fields, or API behavior.