R help - Jan 2012 - Establishing groups using something other than ifelse()

Hello all,

This is one of those "Is there a better way to do this questions". Say
I have a dataframe (df) with a grouping variable (z). This is my base
data. Now I know that there is a higher order level of grouping that
exist for my group variable. So what I want to do is create a new
column that express that higher order level of grouping based on
values in the sub-group (z  in this case). In the past I have used
ifelse() but this tends to get fairly redundant and messy with a large
amount of sub-groupings (z). I've created a sample dataset below. Can
anyone recommend a better way of achieving what I am currently
achieving with ifelse()? A long series of ifelse statements makes me
think that there is something better for this.

## Dataframe creation
df <- data.frame(x=runif(36, 0, 120),
                       y=runif(36, 0, 120),

z=factor(c("A1","A1","A2","A2","B1","B1","B2","B2","C1","C","C2","C2"))
                       )

## Current method is grouping
df$Big.Group <- with(df, ifelse(df$z=="A1","A",
ifelse(df$z=="A2","A",
ifelse(df$z=="B1", "B", ifelse(df$z=="B2",
"B", "C")))))


So any suggestions? Thanks in advance!

Sam

how bout

levels(df$z)[grep('A',levels(df$z))] <- 'A'
levels(df$z)[grep('B',levels(df$z))] <- 'B'
levels(df$z)[grep('C',levels(df$z))] <- 'C'

does that do what you're wanting?


On Thu, Jan 19, 2012 at 3:05 PM, Sam Albers
<tonightsthenight@gmail.com>wrote:
> Hello all,
>
> This is one of those "Is there a better way to do this
questions". Say
> I have a dataframe (df) with a grouping variable (z). This is my base
> data. Now I know that there is a higher order level of grouping that
> exist for my group variable. So what I want to do is create a new
> column that express that higher order level of grouping based on
> values in the sub-group (z  in this case). In the past I have used
> ifelse() but this tends to get fairly redundant and messy with a large
> amount of sub-groupings (z). I've created a sample dataset below. Can
> anyone recommend a better way of achieving what I am currently
> achieving with ifelse()? A long series of ifelse statements makes me
> think that there is something better for this.
>
> ## Dataframe creation
> df <- data.frame(x=runif(36, 0, 120),
>                       y=runif(36, 0, 120),
>
>
z=factor(c("A1","A1","A2","A2","B1","B1","B2","B2","C1","C","C2","C2"))
>                       )
>
> ## Current method is grouping
> df$Big.Group <- with(df, ifelse(df$z=="A1","A",
ifelse(df$z=="A2","A",
> ifelse(df$z=="B1", "B", ifelse(df$z=="B2",
"B", "C")))))
>
>
> So any suggestions? Thanks in advance!
>
> Sam
>
> ______________________________________________
> R-help@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.
>
	[[alternative HTML version deleted]]

Hi Sam,

Check the examples in

require(car)
?recode

HTH,
Jorge.-


On Thu, Jan 19, 2012 at 6:05 PM, Sam Albers <> wrote:
> Hello all,
>
> This is one of those "Is there a better way to do this
questions". Say
> I have a dataframe (df) with a grouping variable (z). This is my base
> data. Now I know that there is a higher order level of grouping that
> exist for my group variable. So what I want to do is create a new
> column that express that higher order level of grouping based on
> values in the sub-group (z  in this case). In the past I have used
> ifelse() but this tends to get fairly redundant and messy with a large
> amount of sub-groupings (z). I've created a sample dataset below. Can
> anyone recommend a better way of achieving what I am currently
> achieving with ifelse()? A long series of ifelse statements makes me
> think that there is something better for this.
>
> ## Dataframe creation
> df <- data.frame(x=runif(36, 0, 120),
>                       y=runif(36, 0, 120),
>
>
z=factor(c("A1","A1","A2","A2","B1","B1","B2","B2","C1","C","C2","C2"))
>                       )
>
> ## Current method is grouping
> df$Big.Group <- with(df, ifelse(df$z=="A1","A",
ifelse(df$z=="A2","A",
> ifelse(df$z=="B1", "B", ifelse(df$z=="B2",
"B", "C")))))
>
>
> So any suggestions? Thanks in advance!
>
> Sam
>
> ______________________________________________
> R-help@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.
>
	[[alternative HTML version deleted]]

On 12-01-19 6:05 PM, Sam Albers wrote:
> Hello all,
>
> This is one of those "Is there a better way to do this
questions". Say
> I have a dataframe (df) with a grouping variable (z). This is my base
> data. Now I know that there is a higher order level of grouping that
> exist for my group variable. So what I want to do is create a new
> column that express that higher order level of grouping based on
> values in the sub-group (z  in this case). In the past I have used
> ifelse() but this tends to get fairly redundant and messy with a large
> amount of sub-groupings (z). I've created a sample dataset below. Can
> anyone recommend a better way of achieving what I am currently
> achieving with ifelse()? A long series of ifelse statements makes me
> think that there is something better for this.
>
> ## Dataframe creation
> df<- data.frame(x=runif(36, 0, 120),
>                         y=runif(36, 0, 120),
>
>
z=factor(c("A1","A1","A2","A2","B1","B1","B2","B2","C1","C","C2","C2"))
>                         )
>
> ## Current method is grouping
> df$Big.Group<- with(df, ifelse(df$z=="A1","A",
ifelse(df$z=="A2","A",
> ifelse(df$z=="B1", "B", ifelse(df$z=="B2",
"B", "C")))))
>
>
> So any suggestions? Thanks in advance!
Use a named vector, e.g.

newcodes <- c(A1="A", A2="A", ... etc.

Then Big.Group <- newcodes[as.character(z)]

Duncan Murdoch