使用前一个单元格的结果来计算当前单元格中 r 中相同变量的值

Question

I am working with a customer database. I would like to compute the value of a cell in a column called EAD1 by using the computed value of previous cell for the same EAD1 column.

An example below shows the data and expected result for EAD1 for data set called data1

Acc_Num, Record_count, BRD, Account_Balance, EAD1  
100, 1, 0.9, 100, 100
100, 2, 0.9, 100,  90  
100, 3, 0.9, 100,  81  
102, 1, 0.8, 100, 100  
102, 2, 0.8, 100,  80   

For Record_count = 1, EAD1 = Account_Balance  
For Record_count > 1, EAD1 = EAD1 previous cell * BRD

Answer 1

Here's a solution using the dplyr package. I've recreated your data using the tribble function, but you won't need to worry about that, other than I've called the dataframe plain_data .

library(dplyr)
#> 
#> Attaching package: 'dplyr'
#> The following objects are masked from 'package:stats':
#> 
#>     filter, lag
#> The following objects are masked from 'package:base':
#> 
#>     intersect, setdiff, setequal, union

plain_data <- tribble(~Acc_Num, ~Record_count, ~BRD, ~Account_Balance, ~EAD1,
100, 1, 0.9, 100, 100,
100, 2, 0.9, 100,  90,  
100, 3, 0.9, 100,  81,  
102, 1, 0.8, 100, 100,  
102, 2, 0.8, 100,  80)  

Here's the solution with dplyr using the pipe operator. We use mutate to create a new column, called new_column . We use the if_else to make the operation conditional on whether the value for each particular row in Record_count is 1 or not. If the condition is met, new_column takes the value of Account_Balance , if not we use the lag function on the column EAD1 , specifying the lag length to be 1 (denoting the previous row in data frame) and calculating that value with the current row's value of BRD .

calculated_data <- plain_data %>% 
  mutate(new_column = if_else(
    condition = Record_count == 1,
    true = Account_Balance, # What gets calculated when Record_count is 1
    false = lag(EAD1, 1) * BRD) # What gets calculated when Record_count is NOT 1
  )

calculated_data
#> # A tibble: 5 × 6
#>   Acc_Num Record_count   BRD Account_Balance  EAD1 new_column
#>     <dbl>        <dbl> <dbl>           <dbl> <dbl>      <dbl>
#> 1     100            1   0.9             100   100        100
#> 2     100            2   0.9             100    90         90
#> 3     100            3   0.9             100    81         81
#> 4     102            1   0.8             100   100        100
#> 5     102            2   0.8             100    80         80

^{Created on 2022-08-12 by the reprex package (v2.0.1)}

Hope this helps. If not, let me know:)