How to mutate without overwriting existing data in data frame?-CodePudding

Below is the example data frame that I will be using. The difference column stands for price difference in time. For example, the price prediction in 2001 will be 6 3 = 9

time	difference	price
2000	NA	6
2001	3	NA
2002	4	NA
2003	6	NA
2004	-8	NA

Right now the code that I am using is below and it is not working since if I were to loop it, it would overwrite previous data when even I do so.

Data <- Data %>%
  mutate(new_price = difference   lag(new_price))

What I am looking for the a piece of code that can finish the list in one click, such that the expecting result would be

time	difference	price
2000	NA	6
2001	3	9
2002	4	13
2003	6	19
2004	-8	11

Any solution is welcomed and thank you guys very much.

CodePudding user response：

Please try this

library(tidyverse)
dat <- data.frame(time=c(2000,2001,2002,2003,2004,2005), diff=c(NA, 3, 4, 6, -8, 10), price=c(6, NA, NA, NA, NA, NA))
dat2 <- dat %>% mutate(price=cumsum(coalesce(diff,price)))

CodePudding user response：

Another approach is to use accumulate from purrr. You provide the vector of data (diff without the first element), the function in this case, and an initial value which is the first value in price.

library(tidyverse)

dat %>%
  mutate(price = accumulate(diff[-1],
                            ` `,
                            .init = price[1]))

Output

  time diff price
1 2000   NA     6
2 2001    3     9
3 2002    4    13
4 2003    6    19
5 2004   -8    11
6 2005   10    21