English | 日本語

dashboardapi

dashboardapi is an unofficial R package for accessing the Statistics Dashboard Web API provided by the Statistics Bureau of Japan. The API requires no registration or API key.

The main workflow follows the same idea as bojapi and WDI: search metadata for an indicator code, find any needed region codes, and retrieve observations as a tidy data frame.

Features

• dashboard_search() / dashboard_indicators(): find indicator codes and inspect cycle, region, seasonal-adjustment, unit, source, and coverage
• dashboard_regions(): find Japanese municipality or ISO country codes
• dashboard_data(): retrieve normalized long or wide observations
• dashboard_terms(), dashboard_events(), and dashboard_surveys(): access the API's auxiliary metadata
• dashboard_codes(): inspect common API enumerations offline
• Automatic batching beyond the API limits of five indicators and 50 regions
• English and Japanese API responses, retries, timeouts, and rate-conscious pauses between automatically batched requests
• Typed API, HTTP, parsing, and no-data conditions

Installation

Install the development version from GitHub:

install.packages("remotes")
remotes::install_github("kenjimyzk/dashboardapi")

To install from a local clone of the repository:

remotes::install_local(".")

Quick start

library(dashboardapi)

# 1. Find an indicator
population_meta <- dashboard_search(
  "Total population (Both sexes)",
  lang = "en"
)

population_meta[, c(
  "indicator_code", "name", "cycle_name",
  "regional_rank_name", "unit"
)]

# 2. Inspect region codes
prefectures <- dashboard_regions(
  parent_region_code = "00000",
  lang = "en"
)

# 3. Retrieve annual Japanese population
population <- dashboard_data(
  indicator_code = c(population = "0201010000000010000"),
  region_code = "00000",
  time_from = "2020CY00",
  time_to = "2024CY00",
  cycle = "year",
  regional_rank = "japan",
  seasonal = "original",
  lang = "en"
)

The time column preserves the API's exact period code. The date column contains the first day of the represented period for analysis: 20240100 becomes 2024-01-01, 20242Q00 becomes 2024-04-01, 2024CY00 becomes 2024-01-01, and 2024FY00 becomes 2024-04-01.

Prefecture example

prefecture_population <- dashboard_data(
  indicator_code = c(population = "0201010000000010000"),
  region_code = c("13000", "27000"),
  time_from = "2020CY00",
  time_to = "2024CY00",
  cycle = "year",
  regional_rank = "prefecture",
  seasonal = "original",
  wide = TRUE
)

Request rate and errors

The official API page asks users not to generate a large access volume in a short period. A request with more than five indicators or 50 regions is split automatically, and dashboardapi waits at least one second between those requests.

options(
  dashboardapi.wait = 2,
  dashboardapi.timeout = 60,
  dashboardapi.retries = 3
)

API errors have class dashboard_api_response_error, communication errors have class dashboard_http_error, and unexpected structures have class dashboard_parse_error. A successful no-data response issues a dashboard_no_data_warning and returns a typed empty tibble.

Attribution and credit for public services

The official Copyright Policy distinguishes three requirements when content from the Statistics Dashboard is published:

1. Cite the source.
2. If content or data are edited, say so and identify the editing entity.
3. If a service uses the API, display the specified API credit.

dashboard_api_credit() returns the corresponding source citation, edited-content template, and API credit. Replace <entity> before publishing edited content:

attribution <- dashboard_api_credit("en")
attribution$source
attribution$processed
attribution$credit

dashboardapi is independently developed and is not affiliated with or endorsed by the Statistics Bureau of Japan. Its MIT license does not relicense government or third-party data, metadata, code systems, or official texts. Check any individual rights statement before reusing third-party content.

Official documentation

• Statistics Dashboard API
• Statistics Dashboard site policy