loop for list element with datetime in r

问题

loop for list element with datetime in r

I have a df with name mistake. I splitted the mistake df by ID. Now I have over 300 different objects in the list.

library(dplyr)
df <- split.data.frame(mistake, mistake$ID)

Every list object has two different datetime stamps. At first I need the minutes between this two datetime stamps. Then I duplicate the rows of the object by the variable stay (this is the difftime between the sat and end time too). Then I overwrite the test variable with the increment n_mintes.

library(lubridate)
start_date <- df[[1]]$datetime
end_date <- df[[1]]$gehtzeit
n_minutes <- interval(start_date,end_date)/minutes(1)
see <- start_date + minutes(0:n_minutes)#the diff time in minutes I need 
df[[1]]$test<- Sys.time()#a new variable 
df[[1]] <- data.frame(df[[1]][rep(seq_len(dim(df[[1]])[1]),df[[1]]$stay+1),1:17, drop= F], row.names=NULL)
df[[1]]$test <- format(start_date + minutes(0:n_minutes), format = "%d.%m.%Y %H:%M:%S")

I want to do this with every objcet of the list. And then 'rbind' or 'unsplit' my list. I know I need a loop. But I don' t know how to do this with the list element. Any help would be create!

Here is a small df example;

mistake


Baureihe Verbund      Fahrzeug Code                           Codetext    Subsystem          Kommt.Zeit
71      411 ICE1166 93805411866-7 1A50 Querfederdruck 1 ungleich Sollwert Neigetechnik 29.07.2018 23:00:07
72      411 ICE1166 93805411866-7 1A50 Querfederdruck 1 ungleich Sollwert Neigetechnik 04.08.2018 11:16:41
             Geht.Zeit           Anstehdauer Jahr Monat   KW Tag Wartung.geht            datetime            gehtzeit
71 29.07.2018 23:02:56 00 Std 02 Min 49 Sek  2018     7 KW30  29            0 2018-07-29 23:00:00 2018-07-29 23:02:00
72 04.08.2018 11:19:20 00 Std 02 Min 39 Sek  2018     8 KW31   4            0 2018-08-04 11:16:00 2018-08-04 11:19:00
   bleiben                                           ID
71  2 secs 2018-07-29 23:00:00 2018-07-29 23:02:00 1A50
72  3 secs 2018-08-04 11:16:00 2018-08-04 11:19:00 1A50

And here ist the structure:

    str(mistake)
'data.frame':   2 obs. of  18 variables:
 $ Baureihe    : int  411 411
 $ Verbund     : Factor w/ 1 level "ICE1166": 1 1
 $ Fahrzeug    : Factor w/ 7 levels "93805411066-4",..: 7 7
 $ Code        : Factor w/ 6 levels "1A07","1A0E",..: 3 3
 $ Codetext    : Factor w/ 6 levels "ITD Karte gestört",..: 5 5
 $ Subsystem   : Factor w/ 1 level "Neigetechnik": 1 1
 $ Kommt.Zeit  : Factor w/ 70 levels "02.08.2018 00:07:23",..: 68 6
 $ Geht.Zeit   : Factor w/ 68 levels "01.08.2018 01:30:25",..: 68 8
 $ Anstehdauer : Factor w/ 46 levels "00 Std 00 Min 01 Sek ",..: 12 4
 $ Jahr        : int  2018 2018
 $ Monat       : int  7 8
 $ KW          : Factor w/ 5 levels "KW27","KW28",..: 4 5
 $ Tag         : int  29 4
 $ Wartung.geht: int  0 0
 $ datetime    : POSIXlt, format: "2018-07-29 23:00:00" "2018-08-04 11:16:00"
 $ gehtzeit    : POSIXlt, format: "2018-07-29 23:02:00" "2018-08-04 11:19:00"
 $ bleiben     :Class 'difftime'  atomic [1:2] 2 3
  .. ..- attr(*, "units")= chr "secs"
 $ ID          : chr  "2018-07-29 23:00:00 2018-07-29 23:02:00 1A50" "2018-08-04 11:16:00 2018-08-04 11:19:00 1A50"

回答1:

Consider building a generalized user-defined function that receives a data frame as input parameter. Then, call the function with by. Like split, by also subsets a data frame by one or more factor(s) such as ID but, unlike split, by can then pass subsets into a function. To row bind all together, run do.call at end.

Below removes the redundant df$test <- Sys.time() which is overwritten later and uses see object inside format() call at end to avoid re-calculation and repetition.

calc_datetime <- function(df) {
   # INITIAL CALCS
   start_date <- df$datetime
   end_date <- df$gehtzeit
   n_minutes <- interval(start_date, end_date)/minutes(1)
   see <- start_date + minutes(0:n_minutes) # the diff time in minutes I need 

   # BUILD OUTPUT DF
   df <- data.frame(df[rep(seq_len(dim(df)[1]), df$stay+1), 1:17, drop= F], row.names=NULL)
   df$test <- format(see, format = "%d.%m.%Y %H:%M:%S")

   return(df)
}

# BUILD LIST OF SUBSETTED DFs
df_list <- by(mistake, mistake$ID, calc_datetime)

# APPEND ALL RESULT DFs TO SINGLE FINAL DF
final_df <- do.call(rbind, df_list)

回答2:

Along the same lines as Parfait's answer, and using the same user defined function calc_datetime, but I would use map_dfr from the purrr package:

df_list <- split(mistake, mistake$ID)
final_df <- map_dfr(df_list, calc_datetime)

If you update the question to have data I can use I can give a demonstration that works

来源：https://stackoverflow.com/questions/51746845/loop-for-list-element-with-datetime-in-r

标签

list

loops

datetime

posixlt