R › R help

by output into data frame

Classic

List

Threaded

4 messages Options

David Perlman

by output into data frame

I could do this in various hacky ways, but what's the right way?

I have a nice application of the by function, which does what I want. The output looks like this:

> auc_stress
lab.samples.stress$subid: 2
cortisol amylase
1 919.05 6834.8
---------------------------------------------------------------------------------------------------------------------------
lab.samples.stress$subid: 3
cortisol amylase
11 728.25 24422.05

etc.

What I want is a data frame roughly like this:

subid cortisol.auc amylase.auc
2 919.05 6834.8
3 728.25 24422.05

etc.

What is a nice way to make that happen?

Here is the code and data that I am using, which should run directly if you copy and paste it:

sanity.check<-read.csv("http://brainimaging.waisman.wisc.edu/~perlman/testdata.csv", header=TRUE, sep = ",")
lab.samples <- subset(sanity.check,Sample!='before bed' & Sample!='morning after')
lab.samples$Sample<-factor(lab.samples$Sample)
lab.samples.stress<-subset(lab.samples,challenge=='stress')
lab.samples.control<-subset(lab.samples,challenge=='control')

auc_ground <- function(sub_df) {
print(sub_df)
auc<-sub_df[1,]*0
timedif<-c(60,10,10,10,10,10,10)
for (i in 1:(nrow(sub_df)-1) ) {
print(c(i,i+1))
#print(c(values[i],values[i+1]))
pair_area<-(sub_df[i,]+sub_df[i+1,])*timedif[i]/2
auc<-auc+pair_area
}
auc
}

auc_stress<-by(lab.samples.stress[c('cortisol','amylase')], lab.samples.stress$subid, auc_ground, simplify=T)
auc_control<-by(lab.samples.control[c('cortisol','amylase')], lab.samples.control$subid, auc_ground, simplify=T)

Thanks for your help!

P.S. sorry if this question has been answered before, it is nearly impossible to get useful google results on search terms like "by"... too common word...

-dave----------------------------------------------------------------------
A neuroscientist is at the video arcade, when someone makes him a $1000 bet
on Pac-Man. He smiles, gets out his screwdriver and takes apart the Pac-Man
game. Everyone says "What are you doing?" The neuroscientist says "Well,
since we all know that Pac-Man is based on electric signals traveling
through these circuits, obviously I can understand it better than the other
guy by going straight to the source!"

______________________________________________
[hidden email] 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.

Jorge I Velez

Re: by output into data frame

Hi David,

Thank you for the reproducible example!

Try

> do.call(rbind, auc_stress)
cortisol amylase
2 919.05 6834.80
3 728.25 24422.05
4 2106.00 25908.35
6 636.40 12209.75
7 1925.95 4749.25

> do.call(rbind, auc_control)
cortisol amylase
2 604.90 2458.00
4 587.65 29954.55
6 493.60 13833.80
7 1211.00 4932.35

HTH,
Jorge.-

On Mon, Mar 19, 2012 at 5:44 PM, David Perlman <> wrote:

> I could do this in various hacky ways, but what's the right way?
>
> I have a nice application of the by function, which does what I want. The
> output looks like this:
>
> > auc_stress
> lab.samples.stress$subid: 2
> cortisol amylase
> 1 919.05 6834.8
>
> ---------------------------------------------------------------------------------------------------------------------------
> lab.samples.stress$subid: 3
> cortisol amylase
> 11 728.25 24422.05
>
> etc.
>
> What I want is a data frame roughly like this:
>
> subid cortisol.auc amylase.auc
> 2 919.05 6834.8
> 3 728.25 24422.05
>
> etc.
>
> What is a nice way to make that happen?
>
>
>
> Here is the code and data that I am using, which should run directly if
> you copy and paste it:
>
>
> sanity.check<-read.csv("
> http://brainimaging.waisman.wisc.edu/~perlman/testdata.csv", header=TRUE,
> sep = ",")
> lab.samples <- subset(sanity.check,Sample!='before bed' & Sample!='morning
> after')
> lab.samples$Sample<-factor(lab.samples$Sample)
> lab.samples.stress<-subset(lab.samples,challenge=='stress')
> lab.samples.control<-subset(lab.samples,challenge=='control')
>
> auc_ground <- function(sub_df) {
> print(sub_df)
> auc<-sub_df[1,]*0
> timedif<-c(60,10,10,10,10,10,10)
> for (i in 1:(nrow(sub_df)-1) ) {
> print(c(i,i+1))
> #print(c(values[i],values[i+1]))
> pair_area<-(sub_df[i,]+sub_df[i+1,])*timedif[i]/2
> auc<-auc+pair_area
> }
> auc
> }
>
> auc_stress<-by(lab.samples.stress[c('cortisol','amylase')],
> lab.samples.stress$subid, auc_ground, simplify=T)
> auc_control<-by(lab.samples.control[c('cortisol','amylase')],
> lab.samples.control$subid, auc_ground, simplify=T)
>
>
> Thanks for your help!
>
> P.S. sorry if this question has been answered before, it is nearly
> impossible to get useful google results on search terms like "by"... too
> common word...
>
>
> -dave----------------------------------------------------------------------
> A neuroscientist is at the video arcade, when someone makes him a $1000 bet
> on Pac-Man. He smiles, gets out his screwdriver and takes apart the Pac-Man
> game. Everyone says "What are you doing?" The neuroscientist says "Well,
> since we all know that Pac-Man is based on electric signals traveling
> through these circuits, obviously I can understand it better than the other
> guy by going straight to the source!"
>
> ______________________________________________
> [hidden email] 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]]

______________________________________________
[hidden email] 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.

Peter Ehlers

Re: by output into data frame

On 2012-03-19 15:00, Jorge I Velez wrote:

> Hi David,
>
> Thank you for the reproducible example!
>
> Try
>
>> do.call(rbind, auc_stress)
> cortisol amylase
> 2 919.05 6834.80
> 3 728.25 24422.05
> 4 2106.00 25908.35
> 6 636.40 12209.75
> 7 1925.95 4749.25
>
>> do.call(rbind, auc_control)
> cortisol amylase
> 2 604.90 2458.00
> 4 587.65 29954.55
> 6 493.60 13833.80
> 7 1211.00 4932.35
>
> HTH,
> Jorge.-

Or with the plyr package:

library(plyr)
ldply(auc_stress)

Peter Ehlers

>
>
> On Mon, Mar 19, 2012 at 5:44 PM, David Perlman<> wrote:
>
>> I could do this in various hacky ways, but what's the right way?
>>
>> I have a nice application of the by function, which does what I want. The
>> output looks like this:
>>
>>> auc_stress
>> lab.samples.stress$subid: 2
>> cortisol amylase
>> 1 919.05 6834.8
>>
>> ---------------------------------------------------------------------------------------------------------------------------
>> lab.samples.stress$subid: 3
>> cortisol amylase
>> 11 728.25 24422.05
>>
>> etc.
>>
>> What I want is a data frame roughly like this:
>>
>> subid cortisol.auc amylase.auc
>> 2 919.05 6834.8
>> 3 728.25 24422.05
>>
>> etc.
>>
>> What is a nice way to make that happen?
>>
>>
>>
>> Here is the code and data that I am using, which should run directly if
>> you copy and paste it:
>>
>>
>> sanity.check<-read.csv("
>> http://brainimaging.waisman.wisc.edu/~perlman/testdata.csv", header=TRUE,
>> sep = ",")
>> lab.samples<- subset(sanity.check,Sample!='before bed'& Sample!='morning
>> after')
>> lab.samples$Sample<-factor(lab.samples$Sample)
>> lab.samples.stress<-subset(lab.samples,challenge=='stress')
>> lab.samples.control<-subset(lab.samples,challenge=='control')
>>
>> auc_ground<- function(sub_df) {
>> print(sub_df)
>> auc<-sub_df[1,]*0
>> timedif<-c(60,10,10,10,10,10,10)
>> for (i in 1:(nrow(sub_df)-1) ) {
>> print(c(i,i+1))
>> #print(c(values[i],values[i+1]))
>> pair_area<-(sub_df[i,]+sub_df[i+1,])*timedif[i]/2
>> auc<-auc+pair_area
>> }
>> auc
>> }
>>
>> auc_stress<-by(lab.samples.stress[c('cortisol','amylase')],
>> lab.samples.stress$subid, auc_ground, simplify=T)
>> auc_control<-by(lab.samples.control[c('cortisol','amylase')],
>> lab.samples.control$subid, auc_ground, simplify=T)
>>
>>
>> Thanks for your help!
>>
>> P.S. sorry if this question has been answered before, it is nearly
>> impossible to get useful google results on search terms like "by"... too
>> common word...
>>
>>
>> -dave----------------------------------------------------------------------
>> A neuroscientist is at the video arcade, when someone makes him a $1000 bet
>> on Pac-Man. He smiles, gets out his screwdriver and takes apart the Pac-Man
>> game. Everyone says "What are you doing?" The neuroscientist says "Well,
>> since we all know that Pac-Man is based on electric signals traveling
>> through these circuits, obviously I can understand it better than the other
>> guy by going straight to the source!"
>>
>> ______________________________________________
>> [hidden email] 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]]
>
> ______________________________________________
> [hidden email] 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.

______________________________________________
[hidden email] 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.

Peter Meilstrup

Re: by output into data frame

In reply to this post by David Perlman

Thanks for providing a reproducible example.

Using the plyr package you can write your whole computation more compactly:

library(plyr)
library(caTools) #for trapz

auc <- ddply(lab.samples, .(challenge, subid),
function(df) {
df$time <- c(0, seq(60,by=10, len=nrow(df)-1))
summarize(df,
cortisol = trapz(time, cortisol),
amylase = trapz(time, amylase))
})

On Mon, Mar 19, 2012 at 2:44 PM, David Perlman <[hidden email]> wrote: