[R] data manipulation between vector and matrix

arun smartpink111 at yahoo.com
Wed Dec 5 23:16:24 CET 2012


HI,
The option z5 takes care of it.
z5<-t(x-t(mat)) #still faster than ?sweep()
 dim(z5)
[1] 20  2
 identical(sweep(-mat,2,x,"+"),z5)
#[1] TRUE


A.K.





________________________________
From: C W <tmrsg11 at gmail.com>
To: arun <smartpink111 at yahoo.com> 
Sent: Wednesday, December 5, 2012 5:09 PM
Subject: Re: [R] data manipulation between vector and matrix


Hi Arun,
Sorry, I might be a little unclear with my words.

The dimensions are different. This is what I got:
> x-t(mat)
     [,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] [,9] [,10] [,11] [,12] [,13] [,14]
[1,]    0   -1   -2   -3   -4   -5   -6   -7   -8    -9   -10   -11   -12   -13
[2,]  -19  -20  -21  -22  -23  -24  -25  -26  -27   -28   -29   -30   -31   -32
     [,15] [,16] [,17] [,18] [,19] [,20]
[1,]   -14   -15   -16   -17   -18   -19
[2,]   -33   -34   -35   -36   -37   -38
> sweep(-mat, 2, x, "+")
      [,1] [,2]
 [1,]    0  -19
 [2,]   -1  -20
 [3,]   -2  -21
 [4,]   -3  -22
 [5,]   -4  -23
 [6,]   -5  -24
 [7,]   -6  -25
 [8,]   -7  -26
 [9,]   -8  -27
[10,]   -9  -28
[11,]  -10  -29
[12,]  -11  -30
[13,]  -12  -31
[14,]  -13  -32
[15,]  -14  -33
[16,]  -15  -34
[17,]  -16  -35
[18,]  -17  -36
[19,]  -18  -37
[20,]  -19  -38
> dim(x-t(mat))
[1]  2 20
> dim(sweep(-mat, 2, x, "+"))
[1] 20  2

On Wed, Dec 5, 2012 at 4:55 PM, arun <smartpink111 at yahoo.com> wrote:

HI Mike,
>I didn't understand "except the x-t(mat) output is very different than the others".  Are you saying that it needs to be transposed?  BTW, that was z5.
>
>A.K.
>
>
>
>
>
>
>
>________________________________
>From: C W <tmrsg11 at gmail.com>
>To: arun <smartpink111 at yahoo.com>
>Cc: R help <r-help at r-project.org>; Sarah Goslee <sarah.goslee at gmail.com>
>Sent: Wednesday, December 5, 2012 4:47 PM
>
>Subject: Re: [R] data manipulation between vector and matrix
>
>
>Thanks for the benchmark.  I actually wanted to go with the winner, except the x-t(mat) output is very different than the others.
>Mike
>
>
>On Wed, Dec 5, 2012 at 4:40 PM, arun <smartpink111 at yahoo.com> wrote:
>
>Hi,
>>
>>By comparing the different methods:
>>set.seed(5)
>> mat1<-matrix(sample(1:1e6,1e6,replace=TRUE),ncol=10000)
>> set.seed(25)
>> x<-sample(1:1e6,10000,replace=TRUE)
>> system.time(z1<-sweep(-mat1,2,x,"+"))
>>#   user  system elapsed
>> # 0.076   0.000   0.069
>> system.time(z2<-apply(-mat1,1,`+`,x))
>> #  user  system elapsed
>> # 0.036   0.000   0.031
>> system.time(z3<-aaply(-mat1,1,`+`,x))
>>#   user  system elapsed
>>#  1.880   0.000   1.704
>> system.time(z4<- x-t(mat1))  #winner
>>#   user  system elapsed
>> # 0.004   0.000   0.007
>> system.time(z5<- t(x-t(mat1)))
>>#   user  system elapsed
>>#  0.008   0.000   0.009
>>
>>
>>A.K.
>>
>>
>>
>>
>>
>>
>>________________________________
>>From: C W <tmrsg11 at gmail.com>
>>To: arun <smartpink111 at yahoo.com>
>>Cc: R help <r-help at r-project.org>; Sarah Goslee <sarah.goslee at gmail.com>
>>Sent: Wednesday, December 5, 2012 4:11 PM
>>
>>Subject: Re: [R] data manipulation between vector and matrix
>>
>>
>>thanks, I knew about apply, but did not you you can put plus signs with quotes.  That's a cool tricky,
>>Mike
>>
>>
>>On Wed, Dec 5, 2012 at 4:05 PM, arun <smartpink111 at yahoo.com> wrote:
>>
>>HI,
>>>In addition to ?sweep(), you can use
>>>
>>>apply(-mat,1,`+`,x)
>>>
>>>#or
>>>library(plyr)
>>>aaply(-mat,1,"+",x)
>>>
>>>
>>>A.K.
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>----- Original Message -----
>>>From: C W <tmrsg11 at gmail.com>
>>>To: Sarah Goslee <sarah.goslee at gmail.com>
>>>Cc: r-help <r-help at r-project.org>
>>>Sent: Wednesday, December 5, 2012 1:51 PM
>>>Subject: Re: [R] data manipulation between vector and matrix
>>>
>>>Thanks, Sarah.  First time heard about sweep(), it worked just the way I
>>>wanted.
>>>Mike
>>>
>>>On Wed, Dec 5, 2012 at 1:42 PM, Sarah Goslee <sarah.goslee at gmail.com> wrote:
>>>
>>>> Hi,
>>>>
>>>> On Wed, Dec 5, 2012 at 1:30 PM, C W <tmrsg11 at gmail.com> wrote:
>>>> > Dear list,
>>>> > I was curious how to subtract a vector from matrix?
>>>> >
>>>> > Say, I have
>>>> >
>>>> > mat <- matrix(1:40, nrow=20, ncol=2)
>>>> >
>>>> > x <-c(1,2)
>>>>
>>>> Thanks for the actual reproducible example.
>>>>
>>>> > I want,
>>>> >
>>>> > x-mat[1,] and x-mat[2,], and so on... Basically, subtract column elements
>>>> > of x against column elements in mat.
>>>> >
>>>> > But x-mat won't do it.
>>>>
>>>> This will (note the modification to get x - mat):
>>>> > sweep(-mat, 2, x, "+")
>>>>       [,1] [,2]
>>>>  [1,]    0  -19
>>>>  [2,]   -1  -20
>>>>  [3,]   -2  -21
>>>>  [4,]   -3  -22
>>>>  [5,]   -4  -23
>>>> etc.
>>>>
>>>> --
>>>> Sarah Goslee
>>>> http://www.functionaldiversity.org
>>>>
>>>
>>>    [[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