In my dataframe named combined I want to do a cumulative difference for multiple columns in one go. Here is the sample code:

#The column names of variables where i want to do cumdiff is
#stored in a vector

stock_vars<-c("bank_s_l", "bank_u_l", "ofi_s_l", "ofi_u_l",
             "pcs_s_l", "pcs_u_l", "govt_s_l","govt_u_l",
             "rotw_s_l", "rotw_u_l","hh_s_l", "hh_u_l",
             "total_s_l", "total_u_l")

#Trying to do cumdiff across columns

combined<-combined%>%mutate_at(stock_vars, ~accumulate(.x-.y))

accumulate did not work for multiple columns. Here is the desired output for first 5 observations of one column

bank_s_l    desired output
-555497.2351    -303573.8477
-251923.3874    -256895.1516
4971.764211     -63187.78958
68159.55379     231826.2753
-163666.7216    

CodePudding user response：

Here is a solution. Write a function, newdiff to compute the differences. It accepts 3 arguments:

1. x the input vector;
2. fill the differences vector has one less element, this argument tells what value will make the output complete;
3. last is the fill value placed last, at the end or at the beginning.

x <- c(-555497.235114064, -251923.387378569, 4971.76421076874, 68159.5537924978, 
       -163666.721553141, 105308.587380903, -398216.766893477, -88543.641285372, 
       -51652.6806000001, -39545.33742, -6600, 72500, -296795, 67737, 
       -54438, -100555, -91767, -26238, -40343, -29728)


newdiff <- function(x, fill = NA, last = TRUE){
  y <- x[-length(x)] - x[-1]
  if(last)
    c(y, fill)
  else
    c(fill, y)
}

res <- cbind(x, y = newdiff(x))
head(res)
#>                x          y
#> [1,] -555497.235 -303573.85
#> [2,] -251923.387 -256895.15
#> [3,]    4971.764  -63187.79
#> [4,]   68159.554  231826.28
#> [5,] -163666.722 -268975.31
#> [6,]  105308.587  503525.35

^{Created on 2022-03-13 by the reprex package (v2.0.1)}

Now to compute the differences of several columns, use a standard mutate(across(.)).

suppressPackageStartupMessages(library(dplyr))

newdiff <- function(x, fill = NA, last = TRUE){
  y <- x[-length(x)] - x[-1]
  if(last)
    c(y, fill)
  else
    c(fill, y)
}

iris %>%
  mutate(across(Sepal.Length:Petal.Width, newdiff)) %>% head()
#>   Sepal.Length Sepal.Width Petal.Length Petal.Width Species
#> 1          0.2         0.5          0.0         0.0  setosa
#> 2          0.2        -0.2          0.1         0.0  setosa
#> 3          0.1         0.1         -0.2         0.0  setosa
#> 4         -0.4        -0.5          0.1         0.0  setosa
#> 5         -0.4        -0.3         -0.3        -0.2  setosa
#> 6          0.8         0.5          0.3         0.1  setosa

^{Created on 2022-03-13 by the reprex package (v2.0.1)}

CodePudding user response：

This is the inverse of cumsum:

my_func <- function(x){
  c( x[1], diff(x,1) ) 
}

library(dplyr)

df %>% 
  mutate(x = my_func(value))

# applying it to iris dataset: (idea from Rui Barradas)

iris %>% 
  mutate(across(Sepal.Length:Petal.Width, my_func)) %>% head()

  Sepal.Length Sepal.Width Petal.Length Petal.Width Species
1          5.1         3.5          1.4         0.2  setosa
2         -0.2        -0.5          0.0         0.0  setosa
3         -0.2         0.2         -0.1         0.0  setosa
4         -0.1        -0.1          0.2         0.0  setosa
5          0.4         0.5         -0.1         0.0  setosa
6          0.4         0.3          0.3         0.2  setosa

     value        x
      <dbl>    <dbl>
 1 -555497. -555497.
 2 -251923.  303574.
 3    4972.  256895.
 4   68160.   63188.
 5 -163667. -231826.
 6  105309.  268975.
 7 -398217. -503525.
 8  -88544.  309673.
 9  -51653.   36891.
10  -39545.   12107.
11   -6600    32945.
12   72500    79100 
13 -296795  -369295 
14   67737   364532 
15  -54438  -122175 
16 -100555   -46117 
17  -91767     8788 
18  -26238    65529 
19  -40343   -14105 
20  -29728    10615 

df <- as_tibble(c(-555497.235114064, -251923.387378569, 4971.76421076874, 68159.5537924978, 
  -163666.721553141, 105308.587380903, -398216.766893477, -88543.641285372, 
  -51652.6806000001, -39545.33742, -6600, 72500, -296795, 67737, 
  -54438, -100555, -91767, -26238, -40343, -29728))