how to generate random postcode in r

Question

I need help in generating an r code that assign the random postcode in a csv file with sample size 5000, sample of file look like as below. 2007, 2008, 2009 and so on are the year

ID  2007    2008    2009    2010    2011    2012    2013    2014    2015    2016    2017
X1                                          
X2                                          
X3                                          
X4                                          
X5                                          
X6                                          
X7                                          
X8                                          
X9                                          
X10 

                                    

I have a separate file where all the postcode saved. Sample of the file of the postcode copied below

BR1 1AA
BR1 1AB
BR1 1AD
BR1 1AE
BR1 1AF
BR1 1AG
BR1 1AH
BR1 1AJ
BR1 1AL
BR1 1AX
BR1 1BA
BR1 1BB
BR1 1BP
BR1 1BQ
BR1 1BS
BR1 1BT
BR1 1BU
BR1 1BW
BR1 1BX
BR1 1BY
BR1 1BZ
BR1 1DA
BR1 1DB
BR1 1DD
BR1 1DE
BR1 1DF
BR1 1DG
BR1 1DH
BR1 1DJ
BR1 1DL
BR1 1DN
BR1 1DP
BR1 1DQ
BR1 1DR
BR1 1DS
BR1 1DT
BR1 1DU
BR1 1DW
BR1 1DX
BR1 1EA
BR1 1EE
BR1 1EG
BR1 1EH
BR1 1EJ
BR1 1EL
BR1 1EN
BR1 1EP
BR1 1ER
BR1 1ES
BR1 1EU
BR1 1EW
BR1 1EX
BR1 1EY
BR1 1EZ
BR1 1GA
BR1 1HA
BR1 1HB
BR1 1HD
BR1 1HE
BR1 1HF
BR1 1HG
BR1 1HH
BR1 1HJ
BR1 1HL
BR1 1HN
BR1 1HP
BR1 1HQ
BR1 1HR
BR1 1HS
BR1 1HT
BR1 1HU
BR1 1HW
BR1 1HX
BR1 1HY
BR1 1HZ
BR1 1JA
BR1 1JB
BR1 1JD
BR1 1JF
BR1 1JG
BR1 1JH
BR1 1JJ
BR1 1JL
BR1 1JN
BR1 1JP
BR1 1JQ
BR1 1JR
BR1 1JS
BR1 1JT
BR1 1JU
BR1 1JW
BR1 1JX
BR1 1JY
BR1 1LA
BR1 1LB
BR1 1LD
BR1 1LE
BR1 1LF
BR1 1LG

I want the distribution of the postcode in data sheet in the following way.Number of postcode lived during 2007 to 2017

%	n
39.7	1985
32.3	1615
15.2	760
6.6	330
3.6	180
1.9	95
0.6	30
0.2	10

In the data sheet there 5000 ids for which I have to fill the postcode for 2007 to 2017. 1985 record should have same postcode during 2007 to 2017 but different from each other.

In second step program pick 1615 postcode and assigned to 1615 records in such a way that during 2007 and 2017 there is one change in postcode ( so they lived on two postcode during study period. And so on.

Answer 1

Defining your input in vectors like so

years <- 2007:2017
target_frequencies <- c(1985L, 1615L, 760L, 330L, 180L, 95L, 30L, 10L)
postcodes <- c("1AA", "1AB", "1AD", "1AE", "1AF", "1AG", "1AH", "1AJ", "1AL", "1AX", "1BA", "1BB", "1BP", "1BQ", "1BS", "1BT", "1BU", "1BW", "1BX", "1BY", "1BZ", "1DA", "1DB", "1DD", "1DE", "1DF", "1DG", "1DH", "1DJ", "1DL", "1DN", "1DP", "1DQ", "1DR", "1DS", "1DT", "1DU", "1DW", "1DX", "1EA", "1EE", "1EG", "1EH", "1EJ", "1EL", "1EN", "1EP", "1ER", "1ES", "1EU", "1EW", "1EX", "1EY", "1EZ", "1GA", "1HA", "1HB", "1HD", "1HE", "1HF", "1HG", "1HH", "1HJ", "1HL", "1HN", "1HP", "1HQ", "1HR", "1HS", "1HT", "1HU", "1HW", "1HX", "1HY", "1HZ", "1JA", "1JB", "1JD", "1JF", "1JG", "1JH", "1JJ", "1JL", "1JN", "1JP", "1JQ", "1JR", "1JS", "1JT", "1JU", "1JW", "1JX", "1JY", "1LA", "1LB", "1LD", "1LE", "1LF", "1LG")

I would approach this with purrr :

library(purrr)

We can define a helper function to generate random postcodes, taking the count of unique postcodes as paramater:

generate_postcodes <- function(count) {
  years_with_new_code <- sort(sample(tail(years, -1), count - 1))
  sample(postcodes)[findInterval(years, years_with_new_code) + 1] %>%
    set_names(years)
}

Testing the helper function

generate_postcodes(2)

# 2007  2008  2009  2010  2011  2012  2013  2014  2015  2016  2017 
# "1HW" "1HW" "1HW" "1HW" "1HW" "1HW" "1HW" "1HW" "1EX" "1EX" "1EX" 

generate_postcodes(6)

# 2007  2008  2009  2010  2011  2012  2013  2014  2015  2016  2017 
# "1JD" "1JD" "1JU" "1HQ" "1HQ" "1HQ" "1JA" "1GA" "1GA" "1EE" "1EE" 

Finally, we can call

imap_dfr(target_frequencies, function(count, code_count) {
  map(seq(count), ~ generate_postcodes(code_count))
}) %>%
.[sample(nrow(.)), ]

returning a randomly ordered tibble with the required properties:

# A tibble: 5,005 x 11
   `2007` `2008` `2009` `2010` `2011` `2012` `2013` `2014` `2015` `2016` `2017`
   <chr>  <chr>  <chr>  <chr>  <chr>  <chr>  <chr>  <chr>  <chr>  <chr>  <chr> 
 1 1HW    1EZ    1EZ    1EZ    1EZ    1EZ    1EZ    1EZ    1EZ    1EZ    1EZ   
 2 1LG    1DG    1DU    1HZ    1HZ    1HZ    1HZ    1HZ    1HZ    1HZ    1HZ   
 3 1HD    1HD    1HD    1HD    1HD    1HD    1HD    1HD    1HD    1HD    1HD   
 4 1HF    1HF    1HF    1HF    1HF    1HF    1HF    1HF    1HF    1HF    1GA   
 5 1JU    1JU    1BP    1BP    1BP    1BP    1BP    1BP    1BP    1BP    1DP   
 6 1EG    1ER    1ER    1ER    1ER    1ER    1ER    1ER    1ER    1ER    1ER   
 7 1EL    1EL    1EL    1EL    1EL    1EL    1EL    1EL    1EL    1EL    1EL   
 8 1DN    1DN    1DN    1DN    1JG    1JG    1JG    1JG    1JG    1JG    1JG   
 9 1HG    1HG    1HG    1HG    1HG    1HG    1HG    1BQ    1BQ    1BQ    1BQ   
10 1ER    1ER    1ER    1ER    1ER    1DW    1DW    1DW    1DW    1JQ    1JQ   
# … with 4,995 more rows