有條件地在R中執行difftime

Question

我有一個數據表，其中包含“每年”檢查的設備。 當設備的檢驗代碼為“ AC”（inspection_disposition_cd =“ AC”）時，我需要計算自其上方的行中列出的檢驗記錄以來經過的天數。

我正在使用的代碼可以很好地進行數學運算，但是，我正在努力進行編碼，以便只有具有AC代碼的行才能執行此操作。 誰能提供任何有關如何挑選這些行的建議？

這是工作代碼：

DispNotClear2$First<-c(0,with(DispNotClear2,difftime(insp_dt[2:length(insp_dt)],insp_dt[1], unit="days")))  
DispNotClear2$BETWEEN<-c(0,with(DispNotClear2,diff(insp_dt[1:(length(insp_dt)-1)],unit="days")))}  

另外，這是我的數據表中的30行示例數據：

library(data.table)
SampleData <- setDT(structure(list(record_num = c(12354L, 6764L, 9959L, 94L, 16463L, 
6859L, 80678L, 87555L, 77980L, 2839L, 4785L, 6082L, 28271L, 54L, 
23000L, 2565L, 3507L, 25607L, 106L, 14621L, 33525L, 6335L, 24970L, 
4851L, 77057L, 17247L, 14488L, 2754L, 88945L, 2710L), device_num = c("2P4564", 
"1P27589", "1P9215", "1P32077", "3P2093", "1P29651", "3P13054", 
"1P14559", "3P7242", "1P30282", "1P12286", "1P13149", "3P10127", 
"1P18885", "4P5121", "1P14621", "1P6278", "1P35255", "1P28871", 
"3P2643", "3P12113", "1P31196", "4P2597", "1W5192", "2P5643", 
"3P6750", "3F1928", "1P11978", "1P33505", "1P14572"), year = c(2016, 
2011, 2016, 2010, 2011, 2013, 2014, 2015, 2013, 2012, 2015, 2016, 
2013, 2015, 2010, 2011, 2011, 2015, 2013, 2012, 2016, 2011, 2015, 
2016, 2012, 2016, 2011, 2015, 2010, 2014), inspection_type_cd = c("CATEGORY 1 (1YR)", 
"CATEGORY 1 (1YR)", "CATEGORY 1 (1YR)", "CATEGORY 1 (1YR)", "CATEGORY 1 (1YR)", 
"CATEGORY 1 (1YR)", "CATEGORY 1 (1YR)", "CATEGORY 1 (1YR)", "CATEGORY 1 (1YR)", 
"CATEGORY 1 (1YR)", "CATEGORY 1 (1YR)", "CATEGORY 1 (1YR)", "CATEGORY 1 (1YR)", 
"CATEGORY 1 (1YR)", "CATEGORY 1 (1YR)", "CATEGORY 1 (1YR)", "CATEGORY 1 (1YR)", 
"CATEGORY 1 (1YR)", "CATEGORY 1 (1YR)", "CATEGORY 1 (1YR)", "CATEGORY 1 (1YR)", 
"CATEGORY 1 (1YR)", "CATEGORY 1 (1YR)", "CATEGORY 1 (1YR)", "CATEGORY 1 (1YR)", 
"CATEGORY 1 (1YR)", "CATEGORY 1 (1YR)", "CATEGORY 1 (1YR)", "CATEGORY 1 (1YR)", 
"CATEGORY 1 (1YR)"), inspection_disposition_cd = c("AC", "AC", 
"AC", "AU", "AC", "AC", "AU", "AC", "AC", "AC", "AC", "AU", "AU", 
"AU", "AU", "AU", "RD", "AU", "AU", "AC", "AU", "AC", "AU", "AU", 
"AC", "AU", "AU", "AC", "AU", "AC"), remarks = c("2015 AFFIRMATION OF CORRECTION", 
NA, "2016 AOC", NA, "PENALTY PAID", "2012 AOC", NA, "2015 AFFIRMATION OF CORRECTION", 
"2013 AFFIRMATION OF CORRECTION", "2011 CORRECTION", "2015 AFFIRMATION OF CORRECTION", 
NA, NA, "2015 CAT1", NA, NA, "AFFIRMATION OF CORRECTION FILED 68 DAYS LATE", 
NA, NA, "2012 CORRECTION", "2016 CAT1", "2010 AFFIRMATION OF CORRECTION", 
NA, NA, "2012 CORRECTION", "2016 CAT1", NA, "2015 AFFIRMATION OF CORRECTION", 
NA, "2013 AOC"), insp_dt = structure(c(16911, 15271, 17060, 14784, 
15014, 16009, 16339, 16658, 16044, 15363, 16576, 17151, 16069, 
16444, 14714, 15184, 15237, 16631, 16059, 15415, 16979, 14995, 
16748, 17137, 15644, 16973, 15211, 16636, 14957, 16087), class = "Date"), 
    days_late = c(NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, 
    NA, NA, NA, NA, NA, 68, NA, NA, NA, NA, NA, NA, NA, NA, NA, 
    NA, NA, NA, NA)), .Names = c("record_num", "device_num", 
"year", "inspection_type_cd", "inspection_disposition_cd", "remarks", 
"insp_dt", "days_late"), class = "data.frame", row.names = c(NA, -30L)))

看起來像

    record_num device_num year inspection_type_cd inspection_disposition_cd                                      remarks    insp_dt days_late
 1:      12354     2P4564 2016   CATEGORY 1 (1YR)                        AC               2015 AFFIRMATION OF CORRECTION 2016-04-20        NA
 2:       6764    1P27589 2011   CATEGORY 1 (1YR)                        AC                                           NA 2011-10-24        NA
 3:       9959     1P9215 2016   CATEGORY 1 (1YR)                        AC                                     2016 AOC 2016-09-16        NA
 4:         94    1P32077 2010   CATEGORY 1 (1YR)                        AU                                           NA 2010-06-24        NA
 5:      16463     3P2093 2011   CATEGORY 1 (1YR)                        AC                                 PENALTY PAID 2011-02-09        NA
 6:       6859    1P29651 2013   CATEGORY 1 (1YR)                        AC                                     2012 AOC 2013-10-31        NA
 7:      80678    3P13054 2014   CATEGORY 1 (1YR)                        AU                                           NA 2014-09-26        NA
 8:      87555    1P14559 2015   CATEGORY 1 (1YR)                        AC               2015 AFFIRMATION OF CORRECTION 2015-08-11        NA
 9:      77980     3P7242 2013   CATEGORY 1 (1YR)                        AC               2013 AFFIRMATION OF CORRECTION 2013-12-05        NA
10:       2839    1P30282 2012   CATEGORY 1 (1YR)                        AC                              2011 CORRECTION 2012-01-24        NA
11:       4785    1P12286 2015   CATEGORY 1 (1YR)                        AC               2015 AFFIRMATION OF CORRECTION 2015-05-21        NA
12:       6082    1P13149 2016   CATEGORY 1 (1YR)                        AU                                           NA 2016-12-16        NA
13:      28271    3P10127 2013   CATEGORY 1 (1YR)                        AU                                           NA 2013-12-30        NA
14:         54    1P18885 2015   CATEGORY 1 (1YR)                        AU                                    2015 CAT1 2015-01-09        NA
15:      23000     4P5121 2010   CATEGORY 1 (1YR)                        AU                                           NA 2010-04-15        NA
16:       2565    1P14621 2011   CATEGORY 1 (1YR)                        AU                                           NA 2011-07-29        NA
17:       3507     1P6278 2011   CATEGORY 1 (1YR)                        RD AFFIRMATION OF CORRECTION FILED 68 DAYS LATE 2011-09-20        68
18:      25607    1P35255 2015   CATEGORY 1 (1YR)                        AU                                           NA 2015-07-15        NA
19:        106    1P28871 2013   CATEGORY 1 (1YR)                        AU                                           NA 2013-12-20        NA
20:      14621     3P2643 2012   CATEGORY 1 (1YR)                        AC                              2012 CORRECTION 2012-03-16        NA
21:      33525    3P12113 2016   CATEGORY 1 (1YR)                        AU                                    2016 CAT1 2016-06-27        NA
22:       6335    1P31196 2011   CATEGORY 1 (1YR)                        AC               2010 AFFIRMATION OF CORRECTION 2011-01-21        NA
23:      24970     4P2597 2015   CATEGORY 1 (1YR)                        AU                                           NA 2015-11-09        NA
24:       4851     1W5192 2016   CATEGORY 1 (1YR)                        AU                                           NA 2016-12-02        NA
25:      77057     2P5643 2012   CATEGORY 1 (1YR)                        AC                              2012 CORRECTION 2012-10-31        NA
26:      17247     3P6750 2016   CATEGORY 1 (1YR)                        AU                                    2016 CAT1 2016-06-21        NA
27:      14488     3F1928 2011   CATEGORY 1 (1YR)                        AU                                           NA 2011-08-25        NA
28:       2754    1P11978 2015   CATEGORY 1 (1YR)                        AC               2015 AFFIRMATION OF CORRECTION 2015-07-20        NA
29:      88945    1P33505 2010   CATEGORY 1 (1YR)                        AU                                           NA 2010-12-14        NA
30:       2710    1P14572 2014   CATEGORY 1 (1YR)                        AC                                     2013 AOC 2014-01-17        NA
    record_num device_num year inspection_type_cd inspection_disposition_cd                                      remarks    insp_dt days_late

Answer 1

dplyr和lubridate的替代解決方案：

library(lubridate)
library(dplyr)
SampleData %>%
        filter(inspection_disposition_cd=="AC") %>% 
        mutate(day_passed = today()-insp_dt)

Answer 2

如果我理解正確，則下面的代碼將記錄的上一行與上一行之間的區別作為其中inspection_disposition_cd=="AC"

SampleData[,day_passed:=ifelse(inspection_disposition_cd=="AC",difftime(shift(insp_dt,1),insp_dt,units = "days"),0)]

或者

SampleData[inspection_disposition_cd=="AC",day_passed:=difftime(shift(insp_dt,1),insp_dt,units = "days")]