rdatasetspkg

R Package to Access Datasets from the Rdatasets Archive

https://github.com/vincentarelbundock/rdatasetspkg

Repository

R Package to Access Datasets from the Rdatasets Archive

README.Rmd

---
output: github_document
---



```{r, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "man/figures/README-",
  out.width = "100%"
)
options(width = 300)
```

# Rdatasets package for `R`

The [Rdatasets Archive](https://vincentarelbundock.github.io/Rdatasets/) is a language-agnostic website that hosts thousands of datasets in CSV format. These datasets can be freely downloaded from the website (with documentation), or from any data analysis environment. 

The present page describes a package for `R` that provides a simple interface to search, download, and view documentation for datasets stored in both CSV and Parquet formats.

## Installation

You can install the development version of Rdatasets from [GitHub](https://github.com/vincentarelbundock/Rdatasetspkg) with:

``` r
remotes::install_github("vincentarelbundock/Rdatasets")

# optional dependency: faster downloads and less bandwidth
install.packages("nanoparquet")
```

## `rdsearch()` Search for datasets

Use `rdsearch()` to find datasets by name, package, or title:

```{r search}
library(Rdatasets)
library(tinytable)

# Search all fields (default behavior)
rdsearch(pattern = "iris") |> head()

# Case-insensitive search for datasets about the Titanic
rdsearch(pattern = "(?i)TITANIC", perl = TRUE)[, 1:4]

# Search only in package names
rdsearch(pattern = "ggplot2movies", field = "package") 

# Search only in dataset names
rdsearch(pattern = "iris", field = "dataset")

# Returns no rows
rdsearch(pattern = "bad_name", field = "dataset") 

# Search only in titles
rdsearch(pattern = "Edgar Anderson", field = "title")
```

## `rddata()` Download datasets

Use `rddata()` to download and load datasets:

```{r download}
# Download the famous André-Michel Guerry dataset
guerry <- rddata("Guerry")
head(guerry, 3)

# Download the Titanic data from a specific package
titanic <- rddata("Titanic", "Stat2Data")
head(titanic, 3)
```

## `rdindex()` Browse the full dataset index

Use `rdindex()` to get the complete list of available datasets:

```{r index}
idx <- rdindex()
cat("Total datasets available:", nrow(idx), "\n")

cat("Number of packages:", length(unique(idx$Package)), "\n")

tail(idx, 10)
```

## `rddocs()` View dataset documentation

Use `rddocs()` to open dataset documentation in your browser:

```{r docs, eval=FALSE}
# Open documentation for the iris dataset
rddocs("iris", "datasets")

# Automatic package detection works here too
rddocs("mtcars")
```

## Performance Features

- **Caching**: All functions cache results by default to speed up repeated access and reduce server load
- **Parquet support**: When `nanoparquet` is installed, datasets are downloaded in Parquet format (faster, smaller)
- **CSV fallback**: Automatically falls back to CSV format when `nanoparquet` is not available

You can disable caching behavior:

```{r cache, eval=FALSE}
options(Rdatasets_cache = FALSE)
```

**Note**: Please keep caching enabled (TRUE) as it makes repeated access faster and avoids overloading the Rdatasets server.

## Output Formats

The package supports three output formats that can be set globally:

```{r format}
# Default: data.frame (no additional dependencies)
options(Rdatasets_cache = FALSE)
options(Rdatasets_class = "data.frame")
rddata("iris") |> class()

# Tibble format (requires tibble package)
options(Rdatasets_class = "tibble")
rddata("iris")

# data.table format (requires data.table package)
options(Rdatasets_class = "data.table")
rddata("iris")

options(Rdatasets_cache = TRUE)
options(Rdatasets_class = "data.frame")
```

The format setting applies to all functions that return data (`rddata()`, `rdindex()`, `rdsearch()`).

Owner

GitHub Events

Committers

Last synced: about 1 year ago

Issues and Pull Requests

Last synced: 12 months ago

Packages