CODEWITHSUNDEEP
r
Carlos Vega Melendez
Carlos Vega Melendez

Reputation: 1

How to in R: Adding rows

I have a data frame named df and I need to combine certain rows together. I made a mistake entering the data and some individuals came out without a treatment. I know below that G13B is actually G13S. My data is organized as soon below and I would like to add very specific rows together.

Example

Individual v1   v2  v3 Treatment
G13B       0    4   6     NA
G13S       1    2   1   Control
G34B       0    4   6     NA
G34S       1    2   1   Control

Wanted, where I added G13B rows to G13S rows.

Individual v1   v2  v3 Treatment
G13S       1    6   7   Control
G34S       1    6   7   Control

I attempted to do Rbind, but all it does it add the column above it.

Upvotes: 0

Views: 312

Answers (1)

MrFlick
MrFlick

Reputation: 206232

If you want to collapse rows based on the first three characters of the Individual ID, you could do something with dplyr. First, your sample data

dd<-structure(list(Individual = structure(1:4, .Label = c("G13B", 
"G13S", "G34B", "G34S"), class = "factor"), v1 = c(0L, 1L, 0L, 
1L), v2 = c(4L, 2L, 4L, 2L), v3 = c(6L, 1L, 6L, 1L), Treatment = structure(c(NA, 
1L, NA, 1L), .Label = "Control", class = "factor")), .Names = c("Individual", 
"v1", "v2", "v3", "Treatment"), class = "data.frame", row.names = c(NA, 
-4L))

then you can do

library(dplyr)

dd %>% group_by(IND=substr(Individual,1,3)) %>% 
    summarize(
       Individual=max(as.character(Individual), na.rm=T), 
       v1=sum(v1), 
       v2=sum(v2), 
       v3=sum(v3), 
       Treatment=min(as.character(Treatment), na.rm=T)) %>% 
    ungroup() %>% select(-IND)

which returns

  Individual v1 v2 v3 Treatment
1       G13S  1  6  7   Control
2       G34S  1  6  7   Control

Upvotes: 1

Related Questions