Dear useRs,
Thanks for advice from Joris Meys,
Now will try to think how to make it working for less specyfic case,
to make the problem more general.
Then the result should be displayed for every group between non empty string in
c2
i.e. not only result for:
#mean:
c1 c3 c4 c5
20 Start1 Stop1 Start1-Stop1
25.48585 Start2 Stop2 Start2-Stop2
but also for every one group created by space between two closest strings in
c2, that contains only seriess of Na, NA, NA, separated from time to time by
one string i.e.:
#mean:
c1 c3 c4 c5
20 Start1 Stop1 Start1-Stop1
.. Stop1 Start2 Stop1-Start2
25.48585 Start2 Stop2 Start2-Stop2
i.e.
to rewrite this maybe for another simpler version of command
but also for every one group created by space between two closest strings in
c2, that contains only seriess of Na, NA, NA, separated from time to time by
one string A, NA, NA, NA, NA, B, NA, NA, NA, C, NA,NA,NA,NA,D, NA,NA
i.e.:
#mean:
c1 c3 c4 c5
20 A B A-B
.. B C B-C
25.48585 C D C-D
...................
Looking for more general method (function), grouping between these letters in
c2,
I will now try to study solution proposed by Joris Meys
Thanks for immediate aswer
Kaluza
-----Wiadomo¶æ oryginalna-----
Od: Joris Meys [mailto:jorism...@gmail.com]
Wys³ano: Cz 2010-06-24 15:14
Do: Eugeniusz Ka³u¿a
DW: r-help@r-project.org
Temat: Re: [R] ?to calculate sth for groups defined between points in one
variable (string), / value separating/ spliting variable into groups by i.e.
between start, NA, NA, stop1, start2, NA, stop2
On Thu, Jun 24, 2010 at 1:18 PM, Eugeniusz Kaluza
<eugeniusz.kal...@polsl.pl> wrote:
>
> Dear useRs,
>
> Thanks for any advices
>
> # I do not know where are the examples how to mark groups
> # based on signal occurence in the additional variable: cf. variable c2,
> # How to calculate different calculations for groups defined by (split by
> occurence of c2 characteristic data)
>
>
> #First example of simple data
> #mexample 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
> c0<-rbind( 1, 2 , 3, 4, 5, 6, 7, 8, 9,10,11, 12,13,14,15,16,17
> )
> c0
> c1<-rbind(10, 20 ,30,40, 50,10,60,20,30,40,50, 30,10,
> 0,NA,20,10.3444)
> c1
> c2<-rbind(NA,"Start1",NA,NA,"Stop1",NA,NA,NA,NA,NA,NA,"Start2",NA,NA,NA,NA,"Stop2")
> c2
> C.df<-data.frame(cbind(c0,c1,c2))
> colnames(C.df)<-c("c0","c1","c2")
> C.df
>
> # preparation of form for explaining further needed result (next 3 lines are
> not needed indeed, they are only to explain how to obtain final result
> c3<-rbind(NA,"Start1","Start1","Start1","Start1","Start2","Start2","Start2","Start2","Start2","Start2","Start2","Start2","Start2","Start2","Start2","Start2")
> c4<-rbind(NA, "Stop1", "Stop1", "Stop1", "Stop1", "Stop2", "Stop2", "Stop2",
> "Stop2", "Stop2", "Stop2", "Stop2", "Stop2", "Stop2", "Stop2", "Stop2",
> "Stop2")
> C.df<-data.frame(cbind(c0,c1,c2,c3,c4))
> colnames(C.df)<-c("c0","c1","c2","c3","c4")
> C.df$c5<-paste(C.df$c3,C.df$c4,sep="-")
> C.df
>
Now this is something I don't get. The list "Start2-Stop2" starts way
before Start2, actually at Stop1. Sure that's what you want?
I took the liberty of showing how to get the data between start and
stop for every entry, and how to apply functions to it. If you don't
get the code, look at
?lapply
?apply
?grep
I also adjusted your example, as you caused all variables to be
factors by using the cbind in the data.frame function. Never do this
unless you're really sure you have to. But I can't think of a case
where that would be beneficial...
...
C.df<-data.frame(c0,c1,c2)
C.df
# find positions
Start <- grep("Start",C.df$c2)
Stop <- grep("Stop",C.df$c2)
# create indices
idx <- apply(cbind(Start,Stop),1,function(i) i[1]:i[2])
names(idx) <- paste("Start",1:length(Start),"-Stop",1:length(Start),sep="")
# Apply the function summary and get a list back named by the interval.
out <- lapply(idx,function(i) summary(C.df[i,1:2]))
out
If you really need to start Start2 right after Stop1, you can use a
similar approach.
Cheers
Joris
> # NEEDED RESULTS
> # needed result
> # for Stat1-Stop1: mean(20,30,40,50)
> # for Stat2-Stop2: mean(c(10,60,20,30,40,50,30,10,0,NA,20,10.3444), na.rm=T)
> #mean:
> c1 c3 c4 c5
> 20 Start1 Stop1 Start1-Stop1
> 25.48585 Start2 Stop2 Start2-Stop2
>
> #sum
> # for Stat1-Stop1: sum(20,30,40,50)
> # for Stat2-Stop2: sum(c(10,60,20,30,40,50,30,10,0,NA,20,10.3444), na.rm=T)
> #sum:
> c1 c3 c4 c5
> 140 Start1 Stop1 Start1-Stop1
> 280.3444 Start2 Stop2 Start2-Stop2
>
> # for Stat1-Stop1: max(20,30,40,50)
> # for Stat2-Stop2: max(c(10,60,20,30,40,50,30,10,0,NA,20,10.3444), na.rm=T)
> #max:
> c1 c3 c4 c5
> 50 Start1 Stop1 Start1-Stop1
> 60 Start2 Stop2 Start2-Stop2
>
> # place of max (in Start1-Stop1: 4 th element in gruop Start1-Stop1
> # place of max (in Start1-Stop1: 2 nd element in gruop Start1-Stop1
>
> c0 c3 c4 c5
> 4 Start1 Stop1 Start1-Stop1
> 2 Start2 Stop2 Start2-Stop2
>
>
> Thanks for any suggestion,
> Kaluza
>
> [[alternative HTML version deleted]]
>
> ______________________________________________
> 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.
>
--
Joris Meys
Statistical consultant
Ghent University
Faculty of Bioscience Engineering
Department of Applied mathematics, biometrics and process control
tel : +32 9 264 59 87
joris.m...@ugent.be
-------------------------------
Disclaimer : http://helpdesk.ugent.be/e-maildisclaimer.php
[[alternative HTML version deleted]]
______________________________________________
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.
