R data frame timestamp manipulation

Question

I have the following data.frame

> head(k)
              begin               end            type total_replies  rpm
1 03/22/15 00:00:00 03/22/15 00:05:00   178.19                 0.50 0.10
2              <NA>              <NA>   178.19.6               0.50 0.10
3 03/22/15 00:05:00 03/22/15 00:10:00   178.1963               0.50 0.10
4              <NA>              <NA>   178.19.75              0.50 0.10
5              <NA>              <NA>   12.221.253             0.75 0.15
6 03/22/15 00:10:00 03/22/15 00:15:00   101.254                1.50 0.30

The NA values correspond to the timestamp above them respectively, meaning the timestamps in row 2 should be the same as row 1 and so on

How do I fill the NA values to match the timestamp above them ?

Check out the `na.locf` function from the `zoo` package. – David Arenburg Apr 29 '15 at 19:55 — David Arenburg, Apr 29 '15 at 19:55

score 1 · Answer 1 · answered Apr 29 '15 at 19:41

1

I think something like this should work

for (i in 1:nrow(k)) {if is.na(k$begin[i]) {
k$begin[i] = k$begin[i-1]}}

answered Apr 29 '15 at 19:41

ytk

2,787
4
27
42

Dominik Safaric · Accepted Answer · 2015-04-29T20:20:12.977

0

Basically, you have to do the following:

replace <- function(dataFrame, forField){
    for (i in c(1:nrow(dataFrame))){
        if (is.na(dataFrame[i,forField])){
            dataFrame[i,forField] <- dataFrame[i-1,forField]
        }
    }
    return (dataFrame)
}

and call the function by

replace(yourDataFrame, "begin")

to replace all begin fields. The same goes for end.

Update output:

enter image description here

edited Apr 29 '15 at 20:20

answered Apr 29 '15 at 19:55

Dominik Safaric

38
7

1

OP doesn't *have* to do anything. He *may* do what you suggested. There are many possible solutions, and yours *could* potentially be one of them. – David Arenburg Apr 29 '15 at 20:00
Never mind, I found the problem. – akrun Apr 29 '15 at 20:29
What was the problem if I might know? I guess you didn't assign the newly build data frame? – Dominik Safaric Apr 29 '15 at 20:39

R data frame timestamp manipulation

2 Answers2