Merge to data sets by date, filling in missing dates in 1st table, R

user_14539467 0
Viewed 35

I'd like to merge the two data sets by the date, keeping all the dates and filling in the totals col with NA when the date doesn't match

Data set 1

 dates      A_totals  
 2015-07-09     1
 2015-07-10     1   
 2015-07-12     2    
 2015-07-14     4   
 2015-07-16     0

Data set 2

 dates      B_totals  
 2015-07-09     2
 2015-07-11     5   
 2015-07-13     6    
 2015-07-15     9   
 2015-07-17     1

Desired Output

 dates      A_totals  B_totals 
 2015-07-09     1       2
 2015-07-10     1       NA
 2015-07-11     NA      5
 2015-07-12     2       NA
 2015-07-13     NA      6
 2015-07-14     4       NA
 2015-07-15     NA      9
 2015-07-16     0       NA
 2015-07-17     NA      1
r
2 Answers

With base R using merge

merge(df1, df2, "dates", all=T)
       dates A_totals B_totals
1 2015-07-09        1        2
2 2015-07-10        1       NA
3 2015-07-11       NA        5
4 2015-07-12        2       NA
5 2015-07-13       NA        6
6 2015-07-14        4       NA
7 2015-07-15       NA        9
8 2015-07-16        0       NA
9 2015-07-17       NA        1

Data

df1 <- structure(list(dates = c("2015-07-09", "2015-07-10", "2015-07-12",
"2015-07-14", "2015-07-16"), A_totals = c(1L, 1L, 2L, 4L, 0L)), class = "data.frame", row.names = c(NA,
-5L))

df2 <- structure(list(dates = c("2015-07-09", "2015-07-11", "2015-07-13",
"2015-07-15", "2015-07-17"), B_totals = c(2L, 5L, 6L, 9L, 1L)), class = "data.frame", row.names = c(NA,
-5L))
user_94620950

Solution

In the tidyverse, a simple dplyr::full_join() should suffice:

library(dplyr)

dataset_1 %>%
  full_join(dataset_2, by = "dates") %>%
  arrange(dates)

Results

Given dataset_*s like those reproduced here

dataset_1 <- structure(
  list(
    dates = structure(
      c(16625, 16626, 16628, 16630, 16632),
      class = "Date"
    ),
    A_totals = c(1, 1, 2, 4, 0)
  ),
  row.names = c(NA, -5L),
  class = "data.frame"
)

dataset_2 <- structure(
  list(
    dates = structure(
      c(16625, 16627, 16629, 16631, 16633),
      class = "Date"
    ),
    B_totals = c(2, 5, 6, 9, 1)
  ),
  row.names = c(NA, -5L),
  class = "data.frame"
)

this solution should yield the following data.frame:

       dates A_totals B_totals
1 2015-07-09        1        2
2 2015-07-10        1       NA
3 2015-07-11       NA        5
4 2015-07-12        2       NA
5 2015-07-13       NA        6
6 2015-07-14        4       NA
7 2015-07-15       NA        9
8 2015-07-16        0       NA
9 2015-07-17       NA        1
user_160289660
Related