Panel lag / difference by country

The two panel primitives everyone hand-rolls (and gets subtly wrong when the frame isn't sorted): the value n years back, and the change since then – grouped by iso3c, ordered by year, so country A's 1960 never leaks into country B's first row.

Usage

lag_by_country(data, value, n = 1, suffix = NULL, by = c("year", "row"))

diff_by_country(data, value, n = 1, suffix = NULL, by = c("year", "row"))

Arguments

data

A panel with iso3c and year.

value

The value column (unquoted).

n

Number of years (or, with by = "row", observations) to lag / difference over (default 1).

suffix

Suffix for the new column. Defaults to "_lag" / "_diff" (with n appended when n > 1, e.g. "_lag5").

by

How "earlier" is found. "year" (default) takes the value n years earlier for the same country, and NA where that year is absent, so a gap in the panel can never pass for a one-year change. "row" takes the previous observation, whatever its year – the 3.0.0 behaviour, for a panel that is irregular by design – and warns when the years are not consecutive. A period column whose labels are not years ("pre-war") needs "row".

Value

data with the lagged / differenced column added. Rows come back sorted by iso3c then year: the calculation reads each country's series in time order, so a row-aligned vector held alongside data will no longer line up.

Examples

df <- data.frame(iso3c = "USA", year = 2000:2003, gdp = c(100, 110, 121, 133))
lag_by_country(df, gdp)
#> # A tibble: 4 × 4
#>   iso3c  year   gdp gdp_lag
#>   <chr> <int> <dbl>   <dbl>
#> 1 USA    2000   100      NA
#> 2 USA    2001   110     100
#> 3 USA    2002   121     110
#> 4 USA    2003   133     121
diff_by_country(df, gdp)
#> # A tibble: 4 × 4
#>   iso3c  year   gdp gdp_diff
#>   <chr> <int> <dbl>    <dbl>
#> 1 USA    2000   100       NA
#> 2 USA    2001   110       10
#> 3 USA    2002   121       11
#> 4 USA    2003   133       12