Adding observations in a table and attribute a given observation (join)

Question

Good afternoon,

I'm analyzing the distribution of observations in a given month, for example:

Date Observations

2010-01 10

2010-03 15

2010-05 16

Question: How do I insert the missing dates (2010-02 and 2010-05) in the table (using other table with all the monthly dates) and attribute a 0 as observations.

Thanks in advance.

Answer 1

We convert the 'Date' to Date class, then use complete expand the dataset by getting the sequence of min/max or first , last 'Date' by '1 month' while fill ing the 'Observations' with 0

library(tidyr)
library(dplyr)
df1 %>%
     mutate(Date = as.Date(Date)) %>%
     complete(Date = seq(first(Date), last(Date), by = '1 month'), 
            fill = list(Observations = 0))

---

If there is another dataset with complete 'Date', then the obvious option is a left_join and then replace the NA elements in 'Observations' with 0 because by default if we don't have a match, it will be filled with NA

left_join(df2, df1, by = 'Date') %>%
     mutate(Observations = replace_na(Observations, 0))

NOTE: df2 is the dataset with complete 'Date'

In case, if the 'df2' have other columns as well, we don't need to select those columns

left_join(df2 %>% 
               select(Date), df1) %>%
     mutate(Observations = replace_na(Observations, 0))

---

In base R , we can use merge

transform(merge(df2, df1, by = 'Date', all.x = TRUE),
      Observations = replace(Observations, is.na(Observations), 0))