Skip to contents

Separate function for total, count, median, mean

Usage

psrc_pums_count(
  so,
  stat_var = NULL,
  group_vars = NULL,
  incl_na = TRUE,
  rr = FALSE
)

psrc_pums_sum(so, stat_var, group_vars = NULL, incl_na = TRUE, rr = FALSE)

psrc_pums_median(so, stat_var, group_vars = NULL, incl_na = TRUE, rr = FALSE)

psrc_pums_mean(so, stat_var, group_vars = NULL, incl_na = TRUE, rr = FALSE)

psrc_pums_summary(so, stat_var, group_vars = NULL, incl_na = TRUE, rr = FALSE)

Arguments

so

The srvyr object returned by get_psrc_pums

stat_var

The numeric variable to summarize

group_vars

Factor variable/s for grouping, as an UPPERCASE string element or list

incl_na

option to remove NA from group_vars (if FALSE, the total may not reflect the full dataset)

rr

optional relative reliability column, i.e. coefficient of variation as category levels (breakpoints: .15/.3./.5 -> good/fair/weak/unreliable)

Value

A table with the variable names and labels, summary statistic and margin of error

Author

Michael Jensen

Examples

if (FALSE) {
library(magrittr)
so <- get_psrc_pums(1, 2019, "h", c("HINCP", "TEN"))
rs <- psrc_pums_median(so, "HINCP", "TEN")
}