[R] merge matrix row data
arun
smartpink111 at yahoo.com
Thu Aug 1 06:35:25 CEST 2013
Hi Elaine,
I am not sure how your "original matrix" keeps on changing from the original post. Here, your statement about rownames are 1, 2, 4 (the answer I already provided in the last post) and the matrix you showed looks different. It seems like GID is the first column in the matrix. I requested you to dput() the data to reduce these confusions.
mat1<-as.matrix(read.table(text="
GID D0989 D9820 D5629 D4327 D2134
1 1 0 0 1 0
2 0 1 1 0 0
4 0 0 1 0 0
5 1 1 0 0 0
7 0 1 0 0 1
",sep="",header=TRUE))
IslandA<-c("GID 1", "GID 5")
IslandB<- c("GID 2", "GID 4", "GID 7")t(sapply(c("IslandA","IslandB"),function(x) {x1<- mat1[match(gsub(".*\\s+","",get(x)),mat1[,1]),-1];(!!colSums(x1))*1}))
# D0989 D9820 D5629 D4327 D2134
#IslandA 1 1 0 1 0
#IslandB 0 1 1 0 1
A.K.
________________________________
From: Elaine Kuo <elaine.kuo.tw at gmail.com>
To: arun <smartpink111 at yahoo.com>
Cc: R help <r-help at r-project.org>
Sent: Thursday, August 1, 2013 12:00 AM
Subject: Re: [R] merge matrix row data
Hello arun
Thanks for the answers.
I understand the answer to question 2.
However, about question 1, sorry I did not clarify the question.
1. If the row names are 1, 2, and 4 etc (numbers) instead of GID 1, GID 2, and GID 3,
is there any modification in need for the code ?
The original data looks like below.
Original matrix
GID D0989 D9820 D5629 D4327 D2134
1 1 0 0 1 0
2 0 1 1 0 0
4 0 0 1 0 0
5 1 1 0 0 0
7 0 1 0 0 1
Resulting matrix
D0989 D9820 D5629 D4327 D2134
Island A 1 1 0 1 0
Island B 0 1 1 0 1
Elaine
On Thu, Aug 1, 2013 at 7:15 AM, arun <smartpink111 at yahoo.com> wrote:
Hi Elaine,
>
>In that case:
>Do you have "GID" in the "IslandA" and "IslandB"s?
>
>
>IslandA<-c("GID 1", "GID 5")
>IslandB<- c("GID 2", "GID 4", "GID 7")
>
>If there is no change in the two "Islands", then using the same dataset:
>
>
>mat1<- as.matrix(read.table(text="
>D0989 D9820 D5629 D4327 D2134
>GID_1 1 0 0 1 0
>GID_2 0 1 1 0 0
>GID_4 0 0 1 0 0
>GID_5 1 1 0 0 0
>GID_7 0 1 0 0 1
>",sep="",header=TRUE))
>
>row.names(mat1)<- gsub(".*\\_","",row.names(mat1)) #to replace the "GID_" from the row.names()
> mat1
>
># D0989 D9820 D5629 D4327 D2134
>
>#1 1 0 0 1 0
>#2 0 1 1 0 0
>
>#4 0 0 1 0 0
>
>#5 1 1 0 0 0
>
>#7 0 1 0 0 1
>
> IslandA<-c("GID 1", "GID 5")
> IslandB<- c("GID 2", "GID 4", "GID 7")
>res<-t(sapply(c("IslandA","IslandB"),function(x) {x1<- mat1[match(gsub(".*\\s+","",get(x)),row.names(mat1)),];(!!colSums(x1))*1}))
>
>res
> # D0989 D9820 D5629 D4327 D2134
>#IslandA 1 1 0 1 0
>#IslandB 0 1 1 0 1
>
>
>Regarding the use of "!!colSums()"
>You can check these:
>
> t(sapply(c("IslandA","IslandB"),function(x) {x1<- mat1[match(gsub(".*\\s+","",get(x)),row.names(mat1)),];!colSums(x1)}))
>
># D0989 D9820 D5629 D4327 D2134
>#IslandA FALSE FALSE TRUE FALSE TRUE
>#IslandB TRUE FALSE FALSE TRUE FALSE
>
>t(sapply(c("IslandA","IslandB"),function(x) {x1<- mat1[match(gsub(".*\\s+","",get(x)),row.names(mat1)),];!!colSums(x1)}))
>
># D0989 D9820 D5629 D4327 D2134
>#IslandA TRUE TRUE FALSE TRUE FALSE
>#IslandB FALSE TRUE TRUE FALSE TRUE
>
># "*1" will replace TRUE with 1 and FALSE with 0.
>
>A.K.
>
>
>
>
>
>
>
>________________________________
>From: Elaine Kuo <elaine.kuo.tw at gmail.com>
>To: arun <smartpink111 at yahoo.com>
>Sent: Wednesday, July 31, 2013 6:58 PM
>
>Subject: Re: [R] merge matrix row data
>
>
>
>Dear Arun,
>
>Thank you for the clear explanation.
>The row.names question is a mistyping, for I do not have enough sleep last night.
>
>Two more questions
>
>1. If the row names are 1, 2, and 4 etc (numbers) instead of GID 1, GID 2, and GID 3,
> is there any modification in need for the code ?
>
>2. Please kindly explain the code
> (!!colSums(x1))*1}
>
> It is the critical part to merge the row data.
>
>Thanks again.
>
>Elaine
>
>
>
>On Thu, Aug 1, 2013 at 6:45 AM, arun <smartpink111 at yahoo.com> wrote:
>
>Dear Elaine,
>>
>>I used that line only because you didn't provide the data using dput(). So, I need to either use delimiter "," or just leave a "space" by first joining the "GID" and the "numbers" using "_". I chose the latter as I didn't had that much time to spent by putting "," between each entries. After that, I removed "_" using the ?gsub(). As Bert pointed out, there are many online resources for understanding regular expression.
>>
>>In this particular case, what I did was to single out the "_" in the first pair of quotes, and replace with space in the second pair of quotes " ". Therefore, "GID_1", would become "GID 1", which is what your original dataset looks like.
>>
>>If you type row.names(mat1) on the R console and enter, you will be able to get the output.
>>
>>Hope it helps.
>>Arun
>>
>>
>>
>>
>>
>>
>>
>>
>>________________________________
>>From: Elaine Kuo <elaine.kuo.tw at gmail.com>
>>To: arun <smartpink111 at yahoo.com>
>>Cc: R help <r-help at r-project.org>
>>Sent: Wednesday, July 31, 2013 5:07 PM
>>Subject: Re: [R] merge matrix row data
>>
>>
>>
>>
>>Dear Arun
>>
>>Thank you for the very useful help.
>>However, please kindly explain the code below.
>>row.names(mat1)<- gsub("[_]"," ",row.names(mat1))
>>
>>
>>1. what does "[_]" mean?
>>2. what does " " mean?
>>3. what does row.names(mat1) mean?
>>
>>I checked ?gsub but still did not get the idea.
>>
>>Thank you again
>>
>>Elaine
>>
>>
>>
>>On Wed, Jul 31, 2013 at 9:35 PM, arun <smartpink111 at yahoo.com> wrote:
>>
>>HI,
>>>
>>>Please use ?dput()
>>>mat1<- as.matrix(read.table(text="
>>>
>>>D0989 D9820 D5629 D4327 D2134
>>>GID_1 1 0 0 1 0
>>>GID_2 0 1 1 0 0
>>>GID_4 0 0 1 0 0
>>>GID_5 1 1 0 0 0
>>>GID_7 0 1 0 0 1
>>>",sep="",header=TRUE))
>>>row.names(mat1)<- gsub("[_]"," ",row.names(mat1))
>>>IslandA<-c("GID 1", "GID 5")
>>>IslandB<- c("GID 2", "GID 4", "GID 7")
>>> res<- t(sapply(c("IslandA","IslandB"),function(x) {x1<-mat1[match(get(x),row.names(mat1)),];(!!colSums(x1))*1} ))
>>>
>>> res
>>># D0989 D9820 D5629 D4327 D2134
>>>#IslandA 1 1 0 1 0
>>>#IslandB 0 1 1 0 1
>>>A.K.
>>>
>>>
>>>
>>>
>>>
>>>----- Original Message -----
>>>From: Elaine Kuo <elaine.kuo.tw at gmail.com>
>>>To: "r-help at stat.math.ethz.ch" <r-help at stat.math.ethz.ch>
>>>Cc:
>>>Sent: Wednesday, July 31, 2013 9:03 AM
>>>Subject: [R] merge matrix row data
>>>
>>>Dear list,
>>>
>>>
>>>
>>>I have a matrix showing the species presence-absence on a map.
>>>
>>>Its rows are map locations, represented by GridCellID, such as GID1 and GID
>>>5.
>>>
>>>Its columns are species ID, such as D0989, D9820, and D5629.
>>>
>>>The matrix is as followed.
>>>
>>>
>>>
>>>Now I want to merge the GridCellID according to the map location of each
>>>island.
>>>
>>>For instance, Island A consist of GID 1 and 5. Island B consist of GID 2,
>>>4, and 7.
>>>
>>>In GID 1 and 5, species D0989 are both 1.
>>>
>>>Then I want to merge GID 1 and 5 into Island A, with species D0989 as 1.
>>>
>>>The original matrix and the resulting matrix are listed below.
>>>
>>>Please kindly advise how to code the calculation in R.
>>>
>>>Please do not hesitate to ask if anything is unclear.
>>>
>>>Thank you in advance.
>>>
>>>
>>>
>>>Elaine
>>>
>>>
>>>
>>>Original matrix
>>>
>>> D0989 D9820 D5629 D4327 D2134
>>>
>>>GID 1 1 0 0 1 0
>>>
>>>GID 2 0 1 1 0 0
>>>
>>>GID 4 0 0 1 0 0
>>>
>>>GID 5 1 1 0 0 0
>>>
>>>GID 7 0 1 0 0 1
>>>
>>>
>>>
>>>Resulting matrix
>>>
>>> D0989 D9820 D5629 D4327 D2134
>>>
>>>Island A 1 1 0 1 0
>>>
>>>Island B 0 1 1 0 1
>>>
>>> [[alternative HTML version deleted]]
>>>
>>>______________________________________________
>>>R-help at 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.
>>>
>>>
>>
>
More information about the R-help
mailing list