Getting an error saying argument is of length zero when looping

1. I have a column named budgetv2 which means the budget for each month, Cost = total expenses in a day, Period = Name of the month in Indonesian to distinguish the budget for each month and lastly Final Budget = the column you want to fill in with the code below that doesn’t run

2. Suppose I have a df as below code in df1

3. The code will fill in the Final Budget column which I filled in with NA

   MEDevel.com: Open-source for Healthcare and Education

   Collecting and validating open-source software for healthcare, education, enterprise, development, medical imaging, medical records, and digital pathology.

   Visit Medevel

4. The code that I want is the example. So the first row is January because that row is the first row of January, so the Final Budget will automatically take the value from the budgetv2 column, so the final budget value is 10000

5. Then in the second row because it is no longer in the first row in January, the Final Budget will make calculations like :
   budgetv2(second row) + Final Budget(top/first row) – Cost(top/first row) then the Final Budget value is in the 2nd column is 10000 + 100000 – 1000 = 19000 is the second row value in january

6. continue in the third row, have the same steps as number 3, namely budgetv2(row 3) + Final Budget(row above/row two) – cost (row above row 3.row second) then the calculation becomes like = 10000 + 19000 – 500 = 285000

7. Row 4 has entered a different month, because row 4 is the first row of the February (not January) period, the Final Budget will take the value of budgetv2, which is 500, so the value of the final budget in row 4 is 500

8. Row 5 is the second row in the February Period, so this second row will have the same formula as in steps 5 and 6, namely Final Budget = budgetv2(5th row) + Final Budget (upper row/4th row) – Cost (upper row/4th row) i.e.. 500 + 500 – 200 = 800 is the value for the Final Budget in row 5

I Already tried this code

df1 <- data.frame(Cost = c(1000,500,200,200,100,300,200),
                  budgetv2 = c(10000,10000,10000,500,500,700,700),
                  Period = c("Januari","Januari","Januari","Februari","Februari","Maret","Maret"),
                  Final_Budget = c(NA,NA,NA,NA,NA,NA,NA))

for (i in 1:nrow(df1)) {
  if (i == 1 | df1$Period[i] != df1$Period[i-1]) { #jika row saat ini adalah row pertama di bulan tersebut
    df1$Final_Budget[i] <- df1$budgetv2[i]
  } else {
df1$Final_Budget[i] - df1$budgetv2[i] + df1$Final_Budget[i-1] - df1$Cost[i-1]
  }
}

and it gave me this error:

Error in if (i == 1 | df1$period[i]!= df1$period[i - 1]) : 
  argument is of length zero

my expected result is like this:

>Solution :

You cannot pass an index of zero to df1[...]. In your first if-statement you have:

i == 1 | df1$Period[i] != df1$Period [i-1]

which returns zero for [i-1] in the first step of the for-loop. Only because you have | it doesn’t mean that the second part won’t be evaluated. I added an ifelse that returns FALSE for i==1 and only evaluates the second part where i != 1.

You are also missing the assignment in your else. I believe you meant to have

df1$Final_Budget[i] <- df1$budgetv2[i] + df1$Final_Budget[i-1] - df1$Cost[i-1]

Instead of

df1$Final_Budget[i] - df1$budgetv2[i] + df1$Final_Budget[i-1] - df1$Cost[i-1]

Notice <- instead of -.

This should work for you:

for (i in 1:nrow(df1)) {
  if (i == 1 | ifelse(i==1, FALSE, df1$Period[i] != df1$Period[i-1])) {
    df1$Final_Budget[i] <- df1$budgetv2[i]
  } else {
    df1$Final_Budget[i] <- df1$budgetv2[i] + df1$Final_Budget[i-1] - df1$Cost[i-1]
  }
}

df1
#>   Cost budgetv2   Period Final_Budget
#> 1 1000    10000  Januari        10000
#> 2  500    10000  Januari        19000
#> 3  200    10000  Januari        28500
#> 4  200      500 Februari          500
#> 5  100      500 Februari          800
#> 6  300      700    Maret          700
#> 7  200      700    Maret         1100

^{Created on 2023-03-24 by the reprex package (v2.0.1)}

Above, I tried to apply the least amount of change to your code. See these variations below for better practices.

Separating your conditions:

for (i in 1:nrow(df1)) {
  if (i == 1) {
    df1$Final_Budget[i] <- df1$budgetv2[i]
  } else if (df1$Period[i] != df1$Period[i-1]) {
    df1$Final_Budget[i] <- df1$budgetv2[i]
  } else {
    df1$Final_Budget[i] <- df1$budgetv2[i] + df1$Final_Budget[i-1] - df1$Cost[i-1]
  }
}

or have an ifelse that evaluates conditions in order:

for (i in 1:nrow(df1)) {
  if (ifelse(i==1, TRUE, df1$Period[i] != df1$Period[i-1])) {
    df1$Final_Budget[i] <- df1$budgetv2[i]
  } else {
    df1$Final_Budget[i] <- df1$budgetv2[i] + df1$Final_Budget[i-1] - df1$Cost[i-1]
  }
}

or use || instead of | as suggested in the comments (only evaluate the right hand side where the left hand side is not true):

for (i in 1:nrow(df1)) {
  if (i == 1 || df1$Period[i] != df1$Period[i-1]) {
    df1$Final_Budget[i] <- df1$budgetv2[i]
  } else {
    df1$Final_Budget[i] <- df1$budgetv2[i] + df1$Final_Budget[i-1] - df1$Cost[i-1]
  }
}

Or simply in dplyr:

library(dplyr)

df1 %>% 
  group_by(Period) %>% 
  mutate(Final_Budget = cumsum(budgetv2) - cumsum(lag(Cost, default = 0)))