How to add row which depends on last row using dpylr and lubridate?

Question

Open/Close Minute Price Data. Next Minute open is always equal to previous minute close. Exemplary data set:

dt                   open  close
1998-01-02 09:30:00  100   101
1998-01-02 09:31:00  101   102
...
1998-01-02 15:59:00  105   106

After the last row I want to add another row like this:

dt                   open  close
1998-01-02 09:30:00  100   101
1998-01-02 09:31:00  101   102
...
1998-01-02 15:59:00  105   106
1998-01-02 16:00:00  106   NA

That is, the timestamp is increased by one minute, open is equal to previous minute close and close is NA. My naive approach does not work:

library(lubridate)
library(dplyr)

data <- add_row(data, dt = max("dt") + minute(1), open = close[[n()]])

Any ideas?

Answer 1

First, you should use minutes (create time periods) instead of minute(used to get minute component of a date-time). Second, in add_row you could not refer to columns in data with string or column name as in other dplyr functions.

One way to do it is:

> data <- data %>% add_row(dt = max(.$dt) + minutes(1), open = last(.$close))
> data
# A tibble: 4 x 3
  dt                   open close
  <dttm>              <int> <int>
1 1998-01-02 09:30:00   100   101
2 1998-01-02 09:31:00   101   102
3 1998-01-02 15:59:00   105   106
4 1998-01-02 16:00:00   106    NA

Where the sample data is:

> dput(data)
structure(list(dt = structure(c(883733400, 883733460, 883756740
), class = c("POSIXct", "POSIXt"), tzone = "UTC"), open = c(100L, 
101L, 105L), close = c(101L, 102L, 106L)), row.names = c(NA, 
-3L), class = c("tbl_df", "tbl", "data.frame"))

Answer 2

We could also use bind_rows

library(tidyverse)
data %>% 
  summarise(dt = max(dt) + minutes(1), open = last(close)) %>% 
  bind_rows(data, .)
# A tibble: 4 x 3
#  dt                   open close
#  <dttm>              <int> <int>
#1 1998-01-02 09:30:00   100   101
#2 1998-01-02 09:31:00   101   102
#3 1998-01-02 15:59:00   105   106
#4 1998-01-02 16:00:00   106    NA