CODEWITHSUNDEEP
r
usavili
usavili

Reputation: 455

Append empty row to a dataframe

I am trying to add an empty row to an existing populated Dataframe.
I am using the below procedure for this, but I do not want to see NA values in the empty row, just need some blank row.

Existing data frame: 

abc
site control    test delta   pval  
1 US15376   3.15% 3.2% 1.59% 0.0022

empty matrix: 

empty=matrix(c(rep.int(NA,length(abc))),nrow=1,ncol=length(abc))  
colnames(empty) = colnames(abc)  

rbind(abc, empty)
    site control test delta   pval  
1 US15376   3.15% 3.2% 1.59% 0.0022  
2    NA    NA NA  NA     NA

Can anyone help me with this?

Upvotes: 31

Views: 67087

Answers (5)

ppoyk
ppoyk

Reputation: 43

If you are using a data.table-object from the widely used package data.table, you can use a method of binding a copy of the last row with the original, and turning it into NA's.

dt <- rbind(dt, dt[nrow(dt), ])
dt[nrow(dt), ] <- NA

This can also be extended to insert the row of NA's to any spot:

i <- 5
dt <- rbind(dt[1:i, ], dt[i:nrow(dt), ])
dt[i, ] <- NA

The previously mentioned simpler method of using rbind(dt, NA) does not work when binding data.table-objects. You would need to define a new matrix, data.frame or data.table of row of NA's with the corresponding column names to the original table, but the above solution is easier.

Upvotes: 0

JGuillamon
JGuillamon

Reputation: 131

tibble´s add_row should do the job. And if after that you want to get rid of NAs, one approach is to use dplyr´s mutate_all

library(tidyverse)

abc %>%
  add_row() %>%
  mutate_all(~replace(., is.na(.), ""))

Upvotes: 2

dpelisek
dpelisek

Reputation: 1041

rbind(abc, NA)

as simple as that

Upvotes: 14

ktiu
ktiu

Reputation: 2636

Here is a version vor tidyverse that can be included in a pipeline:

library(tidyverse)

df %>%
  bind_rows(set_names(rep(NA, ncol(.)), colnames(.)))

On a more philosophical note, "appending an empty row" is something one might want to do when working with (other) statistical software, but I can't think of instances where it aligns well with R's way of doing things. In other words, while it can be done in R, I think that if you are facing this problem, this points to more optimizing potential in your R workflow.

Upvotes: 1

Carolina Fagundes Brinholi
Carolina Fagundes Brinholi

Reputation: 734

This one that was in the comments just worked perfectly for adding an empty row in my data frame:

#before

df

  site control    test delta   pval  
1 US15376   3.15% 3.2% 1.59% 0.0022

df[nrow(df)+1,] <- NA

#after

df

   site control test delta   pval  
1 US15376   3.15% 3.2% 1.59% 0.0022  
2    NA      NA    NA   NA     NA

Upvotes: 54

Related Questions