Skip to contents

Get IMF CPI inflation series for a set of countries

imf_inflation.Rd

Downloads the total CPI index (not seasonally adjusted, monthly frequency) from the IMF's CPI dataset via imf.data::get_data(), and derives month-over-month and year-over-year inflation rates for each country.

Usage

imf_inflation(country_codes = c("DOM"), last_n_obs = 60)

Arguments

country_codes

Character vector of IMF/ISO3 country codes to query (e.g. c("DOM", "COL", "MEX")). Default is "DOM".

last_n_obs

Integer. Number of most recent observations to request per country from the API (passed through to imf.data::get_data()'s last_n_obs argument). Default is 60.

Value

A tibble with one row per country-period, containing:

date

Observation date (first of month), parsed from the IMF TIME_PERIOD field via lubridate::ym().

country

IMF/ISO3 country code.

index

CPI index value for that period (renamed from OBS_VALUE).

vm

Month-over-month inflation rate, in percent.

vi

Year-over-year inflation rate, in percent.

Details

The function queries the IMF CPI dataset with fixed filters for:

  • INDEX_TYPE = "CPI" — the CPI index itself (as opposed to, e.g., core or other index types).

  • COICOP_1999 = "_T" — the "all items" total basket, not a COICOP sub-category.

  • TYPE_OF_TRANSFORMATION = "IX" — index values, not percent-change series.

  • FREQUENCY = "M" — monthly data.

SCALE and STATUS attributes are requested from the API but are not currently kept in the output (they're dropped by the final dplyr::select()).

Inflation rates are computed per country (via .by = COUNTRY) as:

  • vm: month-over-month % change (index vs. 1 month prior)

  • vi: year-over-year % change (index vs. 12 months prior)

Because vm/vi rely on lagged values within each country's series, the first observations for a country will be NA unless last_n_obs covers enough history before the period you actually need (at least 13 months of data for vi to be non-NA on the earliest requested date).

Column names are snake_cased at the end via janitor::clean_names().

Examples

if (FALSE) { # \dontrun{
# Latest 12-month inflation rate for a set of LatAm countries,
# ranked from highest to lowest as of March 2026
imf_inflation(
  country_codes = c("COL", "DOM", "MEX", "BRA", "PER",
                     "CHL", "GTM", "URY", "PRY", "CRI"),
  last_n_obs = 24
) |>
  dplyr::filter(date == "2026-03-01") |>
  dplyr::arrange(dplyr::desc(vi))
} # }