updated wrapr let example.R

## see http://johnmackintosh.com/2017-06-01-UPDATED-let-there-be-progress/ for background
#### direct from the expert - code kindly provided via blog comment by John Mount, Win-Vector LLC

library("dplyr")
library("wrapr")



d <- data.frame(
  location = c('AreaZ', 'AreaZ', 'AreaZ', 'AreaW', 'AreaW', 'AreaW'),
  date = structure(c(14761, 14791, 14822, 14853, 14883, 14914), class = "Date"),
  y = c(6, 3, 8, 18, 18, 2),
  stringsAsFactors = FALSE
)


wrapr_median_rows <- function(dframe, datecol, groupvar, colname, nrows) {
  wrapr::let(c(COL= colname,
               DATES= datecol,
               GROUPVAR= groupvar),
             {
               if(!is.null(groupvar)) {
                 dframe <- dframe %>%
                   dplyr::group_by(GROUPVAR)
               }
               dframe %>%
                 dplyr::arrange(DATES)%>%
                 dplyr::filter(row_number(DATES) %in% c(1:nrows)) %>%
                 dplyr::summarize(baseline = median(COL),
                                  mindate = min(DATES),
                                  maxdate = max(DATES))
             }
  )
}

wrapr_median_rows(d,'date',NULL,'y',2)


##   baseline    mindate    maxdate
## 1      4.5 2010-06-01 2010-07-01


wrapr_median_rows(d,'date','location','y',2)


## # A tibble: 2 x 4
##   location baseline    mindate    maxdate
##      <chr>    <dbl>     <date>     <date>
## 1    AreaW     18.0 2010-09-01 2010-10-01
## 2    AreaZ      4.5 2010-06-01 2010-07-01