inflateR 0.2.0

Bundled data now runs to 2025

Refreshed from the World Bank Development Indicators. Every series gains 2025 except US CPI, which the World Bank has not yet published beyond 2024. The US GDP deflator does reach 2025.

Results will change. Two separate reasons:

• Any call that does not name a to_year defaults to the latest year available, which has moved from 2024 to 2025. For example adjust_inflation(12, 1963, "Australia") returned 212.02 and now returns 218.09. Pass an explicit to_year if you need the old answer.
• The World Bank revised history for several GDP deflator series. The largest is CHF, where 42 years move by a median of 2.4% and up to 13.8%. USD moves by about 1.4% across 60 years, and JPY, NOK, NZD and INR move by smaller amounts. CPI series are essentially unrevised.

Brazil now starts at 1990, not 1980

Rebasing to 2020 = 100 pushes Brazil’s hyperinflation years so far below the base that four decimal places cannot represent them, and they were being stored as exactly zero. A zero index divides, so adjust_inflation(100, 1985, 2020, currency = "BRL") returned Inf rather than refusing the request. Those years are now dropped, so the ordinary coverage check rejects them with a message naming the real range. 40 stored values were affected across brl_cpi and brl_gdp_def.

Fixes to data-raw/cpi_data.R

• The script rounded the rebased index to 2 decimal places, but the shipped data carries 4: 1078 of 1644 values had more precision than re-running the script would have produced. Anyone refreshing the data would have silently degraded it, and flattened Brazil’s early years to zero. Now rounds to 4.
• The end year was hardcoded to 2024 in two places. It now derives from the system date, so the script does not need editing at each World Bank release.

Test fixes surfaced by the refresh

• Three “known value” tests hardcoded index constants (5.71, 66.53, 69.9844) and so failed the moment the World Bank revised history. They now read the index from the bundled data, which tests the arithmetic without turning every refresh into a failure that says nothing about the code.

• Three vectorisation tests asserted result[2] == result[1] * 2 on rounded output. With per-element rounding round(2x, 2) need not equal 2 * round(x, 2), so the assertion was unsound and happened to hold only for the old data. They now pass round = NULL; the round argument keeps its own separate tests.

New functionality

• Added inflation_rate() for calculating cumulative or annualised inflation rates between any two years
• Added list_currencies() to show all supported currencies and their data coverage
• Added round parameter to all four adjustment functions (round = 2 default preserves backward compatibility; use round = NULL for full precision)
• All adjustment functions now accept vector amount inputs
• Refactored internal code: extracted shared helpers to eliminate 4x code duplication across functions
• Added comprehensive test suite (80+ tests covering all functions, all 13 currencies, error handling, inverse properties, and vectorisation)
• Updated data-raw/cpi_data.R to cover all 13 currencies and both CPI and GDP deflator series in a single script

inflateR 0.1.3

• Added 5 new currencies: NZD (New Zealand), INR (India), KRW (South Korea), BRL (Brazil), NOK (Norway)
• Bundled CPI and GDP deflator data for all new currencies (World Bank series)
• BRL CPI coverage begins 1980; all other new series cover 1960–2024

inflateR 0.1.2

• Added adjust_real() for GDP deflator-based adjustment (World Bank NY.GDP.DEFL.ZS)
• Bundled GDP deflator data (1960–2024) for all 8 supported currencies
• adjust_real() accepts the same currency codes and country names as adjust_inflation()

inflateR 0.1.0

• Initial release
• Supports inflation adjustment for GBP, AUD, USD, EUR, CAD, JPY, CNY, and CHF
• Bundled CPI data from 1960–2024 for all currencies (CNY from 1986)
• Currency codes and country names both accepted as input (case-insensitive)
• Data sourced from the World Bank Development Indicators (FP.CPI.TOTL), rescaled to 2020 = 100