Aggregate univariate or multivariate fiscal time series from low to high frequency

Aggregate univariate or multivariate fiscal time series from low to high frequency

Usage

faggr(x, conv_type = "mean", target_freq = "year", na_rm = FALSE)

Arguments

x

a tx-boxable object at a high frequency (e.g. monthly or quarterly)

conv_type

match the aggregated value via "first", "last", "sum", "mean". If conv_type == "uhero" then the name of the time series x is compared to the internal variable sum_pattern. For matching series names the aggregation is based on "sum"; for all others it is based on "mean."

target_freq

target frequency "year", "quarter", "month", "week"

na_rm

logical, if TRUE, incomplete periods are aggregated as well. For irregular series, incomplete periods are always aggregated.

Value

aggregated object of the same type as the input

Details

date stamps typically reflect the start of a period, but here the date stamps in the result are shifted 6 months forward relative to calendar dates, so Jul21 becomes Jan22, or 2022-01-01, and 21Q3 becomes 22Q1.

Examples

monthly_data_example |>
  faggr(conv_type = "sum", target_freq = "quarter")
#> # A tibble: 98 × 4
#>    time       VISNS_HI VAPNS_HI VADCNS_HI
#>    <date>        <dbl>    <dbl>     <dbl>
#>  1 2001-04-01    1686.    1797.      487.
#>  2 2001-07-01    1730.    1774.      531.
#>  3 2001-10-01    1683.    1853.      479.
#>  4 2002-01-01    1650.    1835.      486.
#>  5 2002-04-01    1287.    1440.      402.
#>  6 2002-07-01    1558.    1655.      489.
#>  7 2002-10-01    1600.    1773.      460.
#>  8 2003-01-01    1701.    1900.      502.
#>  9 2003-04-01    1594.    1765.      471.
#> 10 2003-07-01    1580.    1730.      491.
#> # ℹ 88 more rows
monthly_data_example |>
  faggr(conv_type = "uhero", target_freq = "quarter")
#> # A tibble: 98 × 4
#>    time       VISNS_HI VAPNS_HI VADCNS_HI
#>    <date>        <dbl>    <dbl>     <dbl>
#>  1 2001-04-01    1686.    1797.      162.
#>  2 2001-07-01    1730.    1774.      177.
#>  3 2001-10-01    1683.    1853.      160.
#>  4 2002-01-01    1650.    1835.      162.
#>  5 2002-04-01    1287.    1440.      134.
#>  6 2002-07-01    1558.    1655.      163.
#>  7 2002-10-01    1600.    1773.      153.
#>  8 2003-01-01    1701.    1900.      167.
#>  9 2003-04-01    1594.    1765.      157.
#> 10 2003-07-01    1580.    1730.      164.
#> # ℹ 88 more rows
# works with a single series too
monthly_data_example |>
  tsbox::ts_long() |>
  tsbox::ts_pick("VISNS_HI") |>
  faggr(conv_type = "uhero", target_freq = "year") |>
  tsbox::ts_plot()