Two Noobie questions

1. I have a list of lm (linear model) objects. Is it possible to  
select,
through subscripts, a particular element (say, the intercept) from  
all the
models? I've tried something like this:

List[[1:length(list)]][1]
All members of the list are similar. My goal is to have a list of the
intercepts and lists of other estimated parameters. Is it better to  
convert
to a matrix? How to do this?

2. Connected to this, how do I convert from a list back to a vector?  
This
problem arose from using "split" to split a vector by a factor, then
selecting a subset of this (ie. length>10), leaving me with subset  
list of
my original. Unsplit(newList, factor) doesn't work, presumably due  
to my
removal of some values. Thoughts?

On Jan 6, 2009, at 1:50 PM, AllenL wrote:

1. I have a list of lm (linear model) objects. Is it possible to  
select,
through subscripts, a particular element (say, the intercept) from  
all the
models? I've tried something like this:

?coef
if your list of models is ml, then perhaps something like this  
partially tested idea:

lapply(ml, function(x) coef(x)[1] )

This is what I get using that formulation an available logistic model:

 > coef(lr.TC_HDL_BMI)[1]

Intercept
-6.132448

List[[1:length(list)]][1]
All members of the list are similar. My goal is to have a list of the
intercepts and lists of other estimated parameters. Is it better to  
convert
to a matrix? How to do this?

2. Connected to this, how do I convert from a list back to a vector?  
This
problem arose from using "split" to split a vector by a factor, then
selecting a subset of this (ie. length>10), leaving me with subset  
list of
my original. Unsplit(newList, factor) doesn't work, presumably due  
to my
removal of some values. Thoughts?

?unlist

 > ll <- list(1,2,3,4)
 > ll

[[1]]
[1] 1

[[2]]
[1] 2

[[3]]
[1] 3

[[4]]
[1] 4

 > unlist(ll)

[1] 1 2 3 4

 > str(unlist(ll))

  num [1:4] 1 2 3 4

 > is.vector(unlist(ll))

[1] TRUE

-- 
David Winsemius

Thanks!
-Allen



-- 
View this message in context:
http://www.nabble.com/Two-Noobie-questions-tp21316554p21316554.html
Sent from the R help mailing list archive at Nabble.com.

Thanks for your help!

I combined the above two to get the following, which seems to work (if
somewhat inelegant):

int.List<-unlist(lapply(lmList, function(x) {coef(x)[1]}),use.names=FALSE)
lmList is my list of lm objects.
-Allen





David Winsemius wrote:

On Jan 6, 2009, at 1:50 PM, AllenL wrote:

1. I have a list of lm (linear model) objects. Is it possible to
select,
through subscripts, a particular element (say, the intercept) from
all the
models? I've tried something like this:

?coef
if your list of models is ml, then perhaps something like this
partially tested idea:

lapply(ml, function(x) coef(x)[1] )

This is what I get using that formulation an available logistic model:

 > coef(lr.TC_HDL_BMI)[1]

Intercept
-6.132448

List[[1:length(list)]][1]
All members of the list are similar. My goal is to have a list of the
intercepts and lists of other estimated parameters. Is it better to
convert
to a matrix? How to do this?

2. Connected to this, how do I convert from a list back to a vector?
This
problem arose from using "split" to split a vector by a factor, then
selecting a subset of this (ie. length>10), leaving me with subset
list of
my original. Unsplit(newList, factor) doesn't work, presumably due
to my
removal of some values. Thoughts?

?unlist

 > ll <- list(1,2,3,4)
 > ll

[[1]]
[1] 1

[[2]]
[1] 2

[[3]]
[1] 3

[[4]]
[1] 4

 > unlist(ll)

[1] 1 2 3 4

 > str(unlist(ll))

  num [1:4] 1 2 3 4

 > is.vector(unlist(ll))

[1] TRUE

-- 
David Winsemius

Thanks!
-Allen



-- 
View this message in context:
http://www.nabble.com/Two-Noobie-questions-tp21316554p21316554.html
Sent from the R help mailing list archive at Nabble.com.

-- 
View this message in context: 
http://www.nabble.com/Two-Noobie-questions-tp21316554p21317630.html
Sent from the R help mailing list archive at Nabble.com.

summary(lm(x~seq(1,length(x),1)))

summary(lm(x~seq(1,length(x),1)))[[4]][[4]]

summary(lm(x~seq(1,length(x),1)))[[4]][[2]]

On Jan 6, 2009, at 1:50 PM, AllenL wrote:

1. I have a list of lm (linear model) objects. Is it possible to 
select, through subscripts, a particular element (say, the intercept) 
from all the models? I've tried something like this:

?coef
if your list of models is ml, then perhaps something like this  
partially tested idea:

lapply(ml, function(x) coef(x)[1] )

This is what I get using that formulation an available logistic model:

 > coef(lr.TC_HDL_BMI)[1]

Intercept
-6.132448

List[[1:length(list)]][1]
All members of the list are similar. My goal is to have a list of the
intercepts and lists of other estimated parameters. Is it better to  
convert
to a matrix? How to do this?

2. Connected to this, how do I convert from a list back to a vector?  
This
problem arose from using "split" to split a vector by a factor, then
selecting a subset of this (ie. length>10), leaving me with subset  
list of
my original. Unsplit(newList, factor) doesn't work, presumably due  
to my
removal of some values. Thoughts?

?unlist

 > ll <- list(1,2,3,4)
 > ll

[[1]]
[1] 1

[[2]]
[1] 2

[[3]]
[1] 3

[[4]]
[1] 4

 > unlist(ll)

[1] 1 2 3 4

 > str(unlist(ll))

  num [1:4] 1 2 3 4

 > is.vector(unlist(ll))

[1] TRUE

-- 
David Winsemius

Thanks!
-Allen



-- 
View this message in context:
http://www.nabble.com/Two-Noobie-questions-tp21316554p21316554.html
Sent from the R help mailing list archive at Nabble.com.

You can also use subscripts to get at things with a bit of playing around.

summary(lm(x~seq(1,length(x),1)))

Call:
lm(formula = x ~ seq(1, length(x), 1))

Residuals:
    Min       1Q   Median       3Q      Max
-40.0961 -15.5289  -0.6489  12.7488  41.0107

Coefficients:
                      Estimate Std. Error t value Pr(>|t|)
(Intercept)          165.259602   1.620906 101.955   <2e-16 ***
seq(1, length(x), 1)  -0.048711   0.005551  -8.775   <2e-16 ***
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 18.19 on 503 degrees of freedom
Multiple R-squared: 0.1328,     Adjusted R-squared: 0.131
F-statistic:    77 on 1 and 503 DF,  p-value: < 2.2e-16

summary(lm(x~seq(1,length(x),1)))[[4]][[4]]

[1] 0.005551145

summary(lm(x~seq(1,length(x),1)))[[4]][[2]]

[1] -0.04871091

-----Original Message-----
From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org] On Behalf Of AllenL
Sent: 06 January 2009 19:48
To: r-help at r-project.org
Subject: Re: [R] Two Noobie questions


Thanks for your help!

I combined the above two to get the following, which seems to work (if somewhat inelegant):

int.List<-unlist(lapply(lmList, function(x) {coef(x)[1]}),use.names=FALSE) lmList is my list of lm objects.
-Allen





David Winsemius wrote:

On Jan 6, 2009, at 1:50 PM, AllenL wrote:

1. I have a list of lm (linear model) objects. Is it possible to
select, through subscripts, a particular element (say, the intercept)
from all the models? I've tried something like this:

?coef
if your list of models is ml, then perhaps something like this
partially tested idea:

lapply(ml, function(x) coef(x)[1] )

This is what I get using that formulation an available logistic model:

coef(lr.TC_HDL_BMI)[1]

Intercept
-6.132448

List[[1:length(list)]][1]
All members of the list are similar. My goal is to have a list of the
intercepts and lists of other estimated parameters. Is it better to
convert
to a matrix? How to do this?

2. Connected to this, how do I convert from a list back to a vector?
This
problem arose from using "split" to split a vector by a factor, then
selecting a subset of this (ie. length>10), leaving me with subset
list of
my original. Unsplit(newList, factor) doesn't work, presumably due
to my
removal of some values. Thoughts?

?unlist

ll <- list(1,2,3,4)
ll

[[1]]
[1] 1

[[2]]
[1] 2

[[3]]
[1] 3

[[4]]
[1] 4

unlist(ll)

[1] 1 2 3 4

str(unlist(ll))

  num [1:4] 1 2 3 4

is.vector(unlist(ll))

[1] TRUE

--
David Winsemius

Thanks!
-Allen



--
View this message in context:
http://www.nabble.com/Two-Noobie-questions-tp21316554p21316554.html
Sent from the R help mailing list archive at Nabble.com.

-- 
View this message in context: http://www.nabble.com/Two-Noobie-questions-tp21316554p21317630.html
Sent from the R help mailing list archive at Nabble.com.

I have a data frame with unequal rows length separated by comma.....I
have to read the data first and then calculate number of comma in each
row...how can I do that

Regards Rahul