R help - Dec 2011 - How to create a loop and then extract values from the list generated by cor.test

Dear all,

I would like to conduct a Pearson's correlation using cor.test
separately for each individual in my data set. Each individual has
nine observations of the measurement variable 'Delta13C'. I can figure
our how to do it manually one 'Individual_ID' at a time, but I cannot
figure out how to automate the process (a loop I am guessing). In
addition, I would like to generate a new data frame that has the each
individual's 'Individual_ID', the pearson correlation
'estimate', and
the 'p.valve'.

The output table would look something like:

Individual_ID, estimate, p.value
WW_08A_02, -0.7670675, 0.01586
WW_08A_03, -0.02320767, 0.9527

Code with sample data:

## Create sample dataset
WW_Wing_SI <-
structure(list(Individual_ID = c("WW_08A_02", "WW_08A_02",
"WW_08A_02",
"WW_08A_02", "WW_08A_02", "WW_08A_02",
"WW_08A_02", "WW_08A_02",
"WW_08A_02", "WW_08A_03", "WW_08A_03",
"WW_08A_03", "WW_08A_03",
"WW_08A_03", "WW_08A_03", "WW_08A_03",
"WW_08A_03", "WW_08A_03"
), FeatherPosition = c(1, 2, 3, 4, 5, 6, 7, 8, 9, 1, 2, 3, 4,
5, 6, 7, 8, 9), Delta13C = c(-18.67, -19.16, -20.38, -20.96,
-21.61, -21.65, -21.31, -20.8, -21.28, -20.06, -20.3, -21.21,
-22.9, -22.87, -21.13, -20.68, -20.58, -20.69)), .Names =
c("Individual_ID",
"FeatherPosition", "Delta13C"), row.names = c(NA, 18L),
class = "data.frame")


## Create vector with unique individual IDs
WW_Wing_Individuals <- unique(WW_Wing_SI$Individual_ID)

## Create temp dataset with for the first individual
temp <- WW_Wing_SI_Spring[ which(WW_Wing_SI_Spring$Individual_ID
=WW_Wing_Individuals[1]), ]

## Create temp2 dateset with results of pearsons product-moment
correlation (for the first individual)
temp2 <- cor.test(temp$Delta13C, temp$FeatherPosition,
method="pearson")

Many Cheers,
Keith

*******************************************************************************************
Keith Larson, PhD Student
Evolutionary Ecology, Lund University
S?lvegatan 37
223 62 Lund Sweden
Phone: +46 (0)46 2229014 Mobile: +46 (0)73 0465016 Fax: +46 (0)46 2224716
Skype: sternacaspia FB: keith.w.larson at gmail.com

Keith Larson wrote on 12/27/2011 03:05:48 PM:
> Dear all,
> 
> I would like to conduct a Pearson's correlation using cor.test
> separately for each individual in my data set. Each individual has
> nine observations of the measurement variable 'Delta13C'. I can
figure
> our how to do it manually one 'Individual_ID' at a time, but I
cannot
> figure out how to automate the process (a loop I am guessing). In
> addition, I would like to generate a new data frame that has the each
> individual's 'Individual_ID', the pearson correlation
'estimate', and
> the 'p.valve'.
> 
> The output table would look something like:
> 
> Individual_ID, estimate, p.value
> WW_08A_02, -0.7670675, 0.01586
> WW_08A_03, -0.02320767, 0.9527
> 
> Code with sample data:
> 
> ## Create sample dataset
> WW_Wing_SI <-
> structure(list(Individual_ID = c("WW_08A_02",
"WW_08A_02", "WW_08A_02",
> "WW_08A_02", "WW_08A_02", "WW_08A_02",
"WW_08A_02", "WW_08A_02",
> "WW_08A_02", "WW_08A_03", "WW_08A_03",
"WW_08A_03", "WW_08A_03",
> "WW_08A_03", "WW_08A_03", "WW_08A_03",
"WW_08A_03", "WW_08A_03"
> ), FeatherPosition = c(1, 2, 3, 4, 5, 6, 7, 8, 9, 1, 2, 3, 4,
> 5, 6, 7, 8, 9), Delta13C = c(-18.67, -19.16, -20.38, -20.96,
> -21.61, -21.65, -21.31, -20.8, -21.28, -20.06, -20.3, -21.21,
> -22.9, -22.87, -21.13, -20.68, -20.58, -20.69)), .Names = 
c("Individual_ID",
> "FeatherPosition", "Delta13C"), row.names = c(NA, 18L),
class =
"data.frame")
> 
> 
> ## Create vector with unique individual IDs
> WW_Wing_Individuals <- unique(WW_Wing_SI$Individual_ID)
> 
> ## Create temp dataset with for the first individual
> temp <- WW_Wing_SI_Spring[ which(WW_Wing_SI_Spring$Individual_ID =>
WW_Wing_Individuals[1]), ]
> 
> ## Create temp2 dateset with results of pearsons product-moment
> correlation (for the first individual)
> temp2 <- cor.test(temp$Delta13C, temp$FeatherPosition,
method="pearson")
> 
> Many Cheers,
> Keith
> 
> 
*******************************************************************************************
> Keith Larson, PhD Student
> Evolutionary Ecology, Lund University
> Sölvegatan 37
> 223 62 Lund Sweden
> Phone: +46 (0)46 2229014 Mobile: +46 (0)73 0465016 Fax: +46 (0)46 
2224716
> Skype: sternacaspia FB: keith.w.larson@gmail.com

# split your data frame according the the individual IDs
wing.list <- split(WW_Wing_SI, WW_Wing_SI$Individual_ID)

# then apply cor.test() with extract to each element of the list
t(sapply(wing.list, function(temp) 
        cor.test(temp$Delta13C, temp$FeatherPosition, 
        method="pearson")[c("estimate",
"p.value")]))

Jean

	[[alternative HTML version deleted]]

keith.w.larson@gmail.com wrote on 12/27/2011 04:06:09 PM:
> Jean,
> 
> Thanks! I am new to the list and R so I am unsure of the protocol for
> replying to a reply. 

In general, cc R-help so that the entire thread of communication is 
available to everyone

> Your code is simple and elegant. I understand
> most of it, but I do not understand what is happening in the
'sapply'
> portion of your code. 

read
?sapply

lapply(X) returns a list of the same length as X, each element of which is 
the result of applying FUN to the corresponding element of X.

sapply(X) is a user-friendly version and wrapper of lapply(X) by default 
returning a vector, matrix ...

> Second, I know I can create a dataframe from the
> results by:
> 
> wing.pearson <- t(sapply(wing.list, function(temp)
>                                     cor.test(temp$Delta13C,
> temp$FeatherPosition,
>                                    
method="pearson")[c("estimate",
> "p.value")]))
> 
> but when I do this I get the Individual_ID's as row labels, not as a
> variable 'Individual_ID' in my new data frame. Any suggestions?

Actually, the results are in a matrix.  Use the data.frame() function to 
convert the matrix to a data frame.  Then add a new column with the IDs if 
you like.

wing.list <- split(WW_Wing_SI, WW_Wing_SI$Individual_ID)
wing.pearson <- data.frame(t(sapply(wing.list, function(temp) 
cor.test(temp$Delta13C,
        temp$FeatherPosition,
method="pearson")[c("estimate",
"p.value")])))
wing.pearson$Individual_ID <- dimnames(wing.pearson)[[1]]
wing.pearson

> Many Thanks!

You're welcome.

Jean

> Keith
> 
> 
> On Tue, Dec 27, 2011 at 10:41 PM, Jean V Adams <jvadams@usgs.gov>
wrote:
> >
> > Keith Larson wrote on 12/27/2011 03:05:48 PM:
> >
> >
> >> Dear all,
> >>
> >> I would like to conduct a Pearson's correlation using cor.test
> >> separately for each individual in my data set. Each individual has
> >> nine observations of the measurement variable 'Delta13C'.
I can
figure
> >> our how to do it manually one 'Individual_ID' at a time,
but I cannot
> >> figure out how to automate the process (a loop I am guessing). In
> >> addition, I would like to generate a new data frame that has the
each
> >> individual's 'Individual_ID', the pearson correlation
'estimate', and
> >> the 'p.valve'.
> >>
> >> The output table would look something like:
> >>
> >> Individual_ID, estimate, p.value
> >> WW_08A_02, -0.7670675, 0.01586
> >> WW_08A_03, -0.02320767, 0.9527
> >>
> >> Code with sample data:
> >>
> >> ## Create sample dataset
> >> WW_Wing_SI <-
> >> structure(list(Individual_ID = c("WW_08A_02",
"WW_08A_02",
"WW_08A_02",
> >> "WW_08A_02", "WW_08A_02",
"WW_08A_02", "WW_08A_02", "WW_08A_02",
> >> "WW_08A_02", "WW_08A_03",
"WW_08A_03", "WW_08A_03", "WW_08A_03",
> >> "WW_08A_03", "WW_08A_03",
"WW_08A_03", "WW_08A_03", "WW_08A_03"
> >> ), FeatherPosition = c(1, 2, 3, 4, 5, 6, 7, 8, 9, 1, 2, 3, 4,
> >> 5, 6, 7, 8, 9), Delta13C = c(-18.67, -19.16, -20.38, -20.96,
> >> -21.61, -21.65, -21.31, -20.8, -21.28, -20.06, -20.3, -21.21,
> >> -22.9, -22.87, -21.13, -20.68, -20.58, -20.69)), .Names >
>> c("Individual_ID",
> >> "FeatherPosition", "Delta13C"), row.names =
c(NA, 18L), class > >> "data.frame")
> >>
> >>
> >> ## Create vector with unique individual IDs
> >> WW_Wing_Individuals <- unique(WW_Wing_SI$Individual_ID)
> >>
> >> ## Create temp dataset with for the first individual
> >> temp <- WW_Wing_SI_Spring[
which(WW_Wing_SI_Spring$Individual_ID => >> WW_Wing_Individuals[1]), ]
> >>
> >> ## Create temp2 dateset with results of pearsons product-moment
> >> correlation (for the first individual)
> >> temp2 <- cor.test(temp$Delta13C, temp$FeatherPosition, 
method="pearson")
> >>
> >> Many Cheers,
> >> Keith
> >>
> >>
> >> 
> >> Keith Larson, PhD Student
> >> Evolutionary Ecology, Lund University
> >> Sölvegatan 37
> >> 223 62 Lund Sweden
> >> Phone: +46 (0)46 2229014 Mobile: +46 (0)73 0465016 Fax: +46 (0)46 
2224716
> >> Skype: sternacaspia FB: keith.w.larson@gmail.com
> >
> >
> > # split your data frame according the the individual IDs
> > wing.list <- split(WW_Wing_SI, WW_Wing_SI$Individual_ID)
> >
> > # then apply cor.test() with extract to each element of the list
> > t(sapply(wing.list, function(temp)
> >         cor.test(temp$Delta13C, temp$FeatherPosition,
> >         method="pearson")[c("estimate",
"p.value")]))
> >
> > Jean
	[[alternative HTML version deleted]]