CODEWITHSUNDEEP
rdatasetna
Henry dubois
Henry dubois

Reputation: 51

create a data set with only the NA rows in R

How can I create a new data set based on NA value.

                      Date Symbol     Close
1      2005-01-03 01:20:00      A 115.87094
2      2005-01-03 01:25:00      A 115.88168
3      2005-01-03 01:30:00      A 115.88168
4      2005-01-03 01:35:00      A        NA
5      2005-01-03 01:40:00     NA 115.87094
6      2005-01-03 01:45:00     NA 115.87094
7      2005-01-03 01:50:00      A 115.86020

i just want to keep the rows with NA in the row

                    Date   Symbol     Close
4      2005-01-03 01:35:00      A        NA
5      2005-01-03 01:40:00     NA 115.87094
6      2005-01-03 01:45:00     NA 115.87094

Upvotes: 1

Views: 601

Answers (2)

William Bell
William Bell

Reputation: 177

I'm relatively a beginner but you could iterate through the rows (not sure about how you might do it with apply or other vectorized functions) using anyNA() on the row and then if it is true keep the row.

EDIT: e.g.

a <- rep(NA, nrow(df))

for (i in 1:nrow(df)) {
  a[i] <- anyNA(df[i, ])
}

newdf <- df[a, ]

Generally style guides recommend against over using loops, but the anyNA() and is.na() functions are super useful so worth knowing anyways.

Upvotes: -1

akrun
akrun

Reputation: 887048

We can use complete.cases

df1[!complete.cases(df1),]
#                 Date Symbol    Close
#4 2005-01-03 01:35:00      A       NA
#5 2005-01-03 01:40:00   <NA> 115.8709
#6 2005-01-03 01:45:00   <NA> 115.8709

Upvotes: 3

Related Questions