Assign date for a given year & month

This accepts a date, but changes the day. Set/degrade/clump all the days within a month/week to the same day.

clump_month_date( date_detailed, day_of_month=15L )
clump_week_date( date_detailed, day_of_week=2L )

Arguments

date_detailed

The Date value containing the desired year and month. The day will be overwritten. Required

day_of_month

The factor label assigned to the missing value. Defaults to 15 (i.e., the middle of the month).

day_of_week

The factor label assigned to the missing value. Defaults to 2 (i.e., Monday).

Value

An array of Date values.

Details

We use this frequently to set/degrade/clump all the days to the middle of their respective month (ie, the 15th day). The midpoint of a month is usually the most appropriate summary location. It makes graphs more intuitive. Using the midpoint of month can also avoid problems with timezones. A date won't get nudged to a neighboring month accidentally.

Note

A stop error will be thrown if date_detailed is not a Date. day_of_month must be bounded by [1, 31], and day_of_week must be bounded by [1, 7]. Be careful that if you set a November date the 31st day, the result will be December 1st. Consequently, we recommend not setting the day to a value after the 28.

The sql equivalent to clump_month_date() is CAST(convert(char(7), GETDATE(), 126) + '-15' AS date). The sql equivalent to clump_week_date() is SELECT DATEADD(wk, DATEDIFF(wk,0,GETDATE()), 0)

See also

These functions are gloves around lubridate::day() and lubridate::wday()

Essentially the add just error-checking and default values.

Author

Will Beasley

Examples

library(OuhscMunge)
detailed <- seq.Date(from=as.Date("2011-04-21"), to=as.Date("2011-07-14"), by="day")

clumped_month <- clump_month_date(detailed)
table(clumped_month)
#> clumped_month
#> 2011-04-15 2011-05-15 2011-06-15 2011-07-15 
#>         10         31         30         14 
# 2011-04-15 2011-05-15 2011-06-15 2011-07-15
#         10         31         30         14

clumped_week <- clump_week_date(detailed)
table(clumped_week)
#> clumped_week
#> 2011-04-18 2011-04-25 2011-05-02 2011-05-09 2011-05-16 2011-05-23 2011-05-30 
#>          3          7          7          7          7          7          7 
#> 2011-06-06 2011-06-13 2011-06-20 2011-06-27 2011-07-04 2011-07-11 
#>          7          7          7          7          7          5 
# 2011-04-18 2011-04-25 2011-05-02 2011-05-09 2011-05-16 2011-05-23 2011-05-30
#          3          7          7          7          7          7          7
# 2011-06-06 2011-06-13 2011-06-20 2011-06-27 2011-07-04 2011-07-11
#          7          7          7          7          7          5