Skip to contents

Year-on-year ratio of pooled shares with Katz confidence interval

Source: R/emerging.R
compute_yoy_ratio.Rd

Compares the pooled share of the last months periods ending at end with the pooled share of the same periods lag periods earlier.

Usage

compute_yoy_ratio(
  panel,
  key_cols,
  time_col = "mese_idx",
  x_col = "x",
  n_col = "n",
  months = 3L,
  lag = 12L,
  end = NULL,
  conf = 0.95
)

Arguments

panel

A share panel with one row per key and period, e.g. from compute_share_panel(). Periods with n = 0 are ignored.

key_cols

Character vector of columns identifying a series (e.g. "skill_id" or c("cp4", "skill_id")).

time_col

Numeric period index column. Default "mese_idx".

x_col

Count column. Default "x".

n_col

Denominator column. Default "n".

months

Number of periods pooled in each block. Default 3.

lag

Distance in periods between the two blocks. Default 12.

end

Last period of the window. Default: the maximum period in panel.

conf

Confidence level of the interval. Default 0.95.

Value

A data.table with key_cols, x_attuale, n_attuale, x_precedente, n_precedente, yoy (ratio), yoy_lo, yoy_hi, log_yoy, se_log and p_value (two-sided test of yoy = 1). Series with a zero denominator in either block get NA.

Details

With \(x_1, n_1\) the counts and denominators summed over the current block and \(x_0, n_0\) over the reference block, the ratio is \(R = (x_1/n_1)/(x_0/n_0)\). The Katz interval uses \(SE(\log R) = \sqrt{1/x_1 - 1/n_1 + 1/x_0 - 1/n_0}\). When either count is zero, 0.5 is added to both counts and both denominators of that series.

Examples

panel <- data.table::CJ(skill_id = "a", mese_idx = 1:15)
panel[, `:=`(n = 1000L, x = ifelse(mese_idx > 12, 60L, 30L))]
#> Key: <skill_id, mese_idx>
#>     skill_id mese_idx     n     x
#>       <char>    <int> <int> <int>
#>  1:        a        1  1000    30
#>  2:        a        2  1000    30
#>  3:        a        3  1000    30
#>  4:        a        4  1000    30
#>  5:        a        5  1000    30
#>  6:        a        6  1000    30
#>  7:        a        7  1000    30
#>  8:        a        8  1000    30
#>  9:        a        9  1000    30
#> 10:        a       10  1000    30
#> 11:        a       11  1000    30
#> 12:        a       12  1000    30
#> 13:        a       13  1000    60
#> 14:        a       14  1000    60
#> 15:        a       15  1000    60
compute_yoy_ratio(panel, key_cols = "skill_id")
#> Key: <skill_id>
#>    skill_id x_attuale n_attuale x_precedente n_precedente   yoy   yoy_lo
#>      <char>     <int>     <int>        <int>        <int> <num>    <num>
#> 1:        a       180      3000           90         3000     2 1.560848
#>     yoy_hi   log_yoy    se_log      p_value
#>      <num>     <num>     <num>        <num>
#> 1: 2.56271 0.6931472 0.1264911 4.257838e-08