[R] Help needed for efficient way to loop through rows and columns

[David L Carlson]

This will accomplish what you want and should be relatively easy to modify.

# Create data.fram
names <- c("S1", "S2", "S3", "S4")
X <- c("BB", "AA", "AB", "AA")
Y <- c("BB", "BB", "AB", "AA")
Z <- c("BB", "BB", "AA", NA)
AorB <- c("A", "A", "B", "B")
sample <- data.frame(names, X, Y, Z, AorB, 
  stringsAsFactors=FALSE)

# Create recoded data.frame
samplemod <- sample
samplemod[,2:4] <- NA

# Recoded values into samplemod
for (i in 1:nrow(sample)) {
  for (j in 2:4) {
    if (!is.na(sample[i,j])){
      if (sample[i, 5] == "A") {
        samplemod[i,j] <- switch(sample[i,j], AA = 2, 
          AB = 1, BA = 1, BB = 0)
      }
      else { 
        if (sample[i, 5] == "B") {
        samplemod[i,j] <- switch(sample[i,j], AA = 0, 
          AB = 1, BA = 1, BB = 2)
        }
      }
    }
  }
}



----------------------------------------------
David L Carlson
Associate Professor of Anthropology
Texas A&M University
College Station, TX 77843-4352



From: Priya Bhatt [mailto:bhattp60 at gmail.com] 
Sent: Wednesday, May 16, 2012 1:55 PM
To: dcarlson at tamu.edu; r-help at r-project.org
Subject: Re: [R] Help needed for efficient way to loop through rows and
columns

Yes here it is.  I actually convert them all as strings, initially using  
options(stringsAsFactors=F) at the top of my code.

This what the initial dataframe looks like.  Please note this is a toy
dataset:

names    X    Y    Z    AorB
S1    BB    BB    BB    A
S2    AA    BB    BB    A
S3    AB    AB    AA    B
S4    AA    AA    NA    B


And the code to create this initial dataframe is:

names <- c("S1", "S2", "S3", "S4")
X <- c("BB", "AA", "AB", "AA")
Y <- c("BB", "BB", "AB", "AA")
Z <- c("BB", "BB", "AA", NA)
AorB <- c("A", "A", "B", "B")

sample <- data.frame(names, X, Y, Z, AorB)


The final data.frame should look like:

names  X    Y    Z    AorB
S1    0    0    0    A
S2    2    0    0    A
S3    1    1    0    B
S4    0    0    NA B

You're right! - I'll should be able to globally change all ABs and BAs to
1s. Thanks:)  I'm not exactly sure how to change AA and BB depending on AorB
for each row though.  Thoughts?

Thanks for your help thus far, David.

Best, Priya

On Wed, May 16, 2012 at 6:53 AM, David L Carlson <dcarlson at tamu.edu> wrote:
Can you show us what you want the final data.frame to look like? You've
created five variables stored as factors and you seem to be trying to change
those to numeric values? Is that correct?

Since AB and BA are always set to 1, you could just replace those values
globally rather than mess with the ifelse commands for those values. Only AA
and BB are affected by the value of AorB.

Your apply() function processes the data.frame by row so i is a vector
consisting of all the values in the row. You seem to be coding as if i was a
single integer (as in a for loop).

----------------------------------------------
David L Carlson
Associate Professor of Anthropology
Texas A&M University
College Station, TX 77843-4352


> -----Original Message-----
> From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-
> project.org] On Behalf Of Priya Bhatt
> Sent: Wednesday, May 16, 2012 3:08 AM
> To: r-help at r-project.org
> Subject: [R] Help needed for efficient way to loop through rows and
> columns
>
> Dear R-helpers:
>
> I am trying to write a script that iterates through a dataframe that
> looks
> like this:
>
>
> Example dataset called "sample":
>
> names <- c("S1", "S2", "S3", "S4")
> X <- c("BB", "AB", "AB", "AA")
> Y <- c("BB", "BB", "AB", "AA")
> Z <- c("BB", "BB", "AB", NA)
> AorB <- c("A", "A", "A", "B")
>
> sample <- data.frame(names, X, Y, Z, AorB)
>
>
> for a given row,
>
> if AorB == A, then AA == 2, AB = 1, BA = 1, BB = 0
>
> if AorB == B, then AA == 0, AB = 1, BA = 1, BB = 2
>
> I've been trying  to write this using apply and ifelse statements in
> hopes
> that my code runs quickly, but I'm afraid I've make a big mess.  See
> below:
>
> apply(sample, 1, function(i) {
>
>
>   ifelse(sample$AorB[i] == "A",
>          (ifelse(sample[i,] == "AA", sample[i,] <- 2 ,
>                  ifelse(sample[i,] == "AB" || sample[i,] == "BA" ,
> sample[i,] <- 1,
>                         ifelse(sample[i,] == "BB", sample[i,] <- 0,
> sample[i,] <- NA )) )
>           )   , ifelse(sample$AorB[i,] == "B"),
>          (ifelse(sample[i,] == "AA", sample[i,] <- 0 ,
>                  ifelse(sample[i,] == "AB" || sample[i,] == "BA" ,
> sample[i,] <- 1,
>                         ifelse(sample[i,] == "BB", sample[i,] <- 2,
> sample[i,] <- NA))))) })
>
>
> Any Advice?
>
>       [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-
> guide.html
> and provide commented, minimal, self-contained, reproducible code.