[R] Initializing vector and matrices

Jeff Newmiller jdnewm|| @end|ng |rom dcn@d@v|@@c@@u@
Sat Mar 2 19:52:50 CET 2024 

To be fair, these replies no longer include the original question, which was IMO really quite clear (if misguided), and was actually targeted at understanding pre-allocation for better performance. Richard's suggestion to store the along-the-way constructed vectors in a list and examine the lengths of the elements of that list after it is completely constructed if it is really desired to force vectors of varying lengths into a matrix seems appropriate. The possibly-misguided part is in the thought that vectors of different lengths belong in a matrix at all... leaving them in a list seems more likely to be appropriate.

OP: Avoid mis-using NA to "fill out" the end of uniform-length vectors. Each index position in a vector should represent some information... if you are throwing NAs or zeros in willy-nilly then you are not taking care to assign meaning to those positions, and likely wasting memory and time doing so. If you were taking such care, then any appropriate NAs would already be present in the vectors before you put them into the matrix, and the original question would not have crossed OPs mind because the vectors world already be of the same length. NA means that information for a specific position is "unknown", not "space filler".

Now if OP wants better guidance on restructuring their analysis, then certainly sharing those details would help the discussion move away from the original misguided question... but that would be another thread to answer the question that should have been asked.

On March 2, 2024 9:53:41 AM PST, Bert Gunter <bgunter.4567 using gmail.com> wrote:
>"It would be really really helpful to have a clearer idea of what you
>are trying to do."
>
>Amen!
>
>But in R, "constructing" objects by extending them piece by piece is
>generally very inefficient (e.g.
>https://r-craft.org/growing-objects-and-loop-memory-pre-allocation/),
>although sometimes?/often? unavoidable (hence the relevance of your
>comment above). R generally prefers to take a "whole object" point of
>view ( see R books by Chambers, et. al.) and provides code for basic
>operations like vector/matrix, etc. construction to do so. When this
>is not possible, I suspect "optimal" efficient strategies for
>allocating space to build objects gets you into the weeds of how R
>works.
>
>Cheers,
>Bert
>
>
>
>
>On Sat, Mar 2, 2024 at 1:02 AM Richard O'Keefe <raoknz using gmail.com> wrote:
>>
>> The matrix equivalent of
>>   x <- ...
>>   v <- ...
>>   x[length(x)+1] <- v
>> is
>>   m <- ...
>>   r <- ...
>>   m <- rbind(m, r)
>> or
>>   m <- ...
>>   k <- ...
>>   m <- cbind(m, c)
>>
>> A vector or matrix so constructed never has "holes" in it.
>> It's better to think of CONSTRUCTING vectors and matrices rather than
>> INITIALISING them,
>> because always being fully defined is important.
>>
>> It would be really really helpful to have a clearer idea of what you
>> are trying to do.
>>
>> On Fri, 1 Mar 2024 at 03:31, Ebert,Timothy Aaron <tebert using ufl.edu> wrote:
>> >
>> > You could declare a matrix much larger than you intend to use. This works with a few megabytes of data. It is not very efficient, so scaling up may become a problem.
>> > m22 <- matrix(NA, 1:600000, ncol=6)
>> >
>> > It does not work to add a new column to the matrix, as in you get an error if you try m22[ , 7] but convert to data frame and add a column
>> >
>> > m23 <- data.frame(m22)
>> > m23$x7 <- 12
>> >
>> > The only penalty that I know of to having unused space in a matrix is the amount of memory it takes. One side effect is that your program may have a mistake that you would normally catch with a subscript out of bounds error but with the extra space it now runs without errors.
>> >
>> > Tim
>> >
>> >
>> >
>> > -----Original Message-----
>> > From: R-help <r-help-bounces using r-project.org> On Behalf Of Richard O'Keefe
>> > Sent: Thursday, February 29, 2024 5:29 AM
>> > To: Steven Yen <styen using ntu.edu.tw>
>> > Cc: R-help Mailing List <r-help using r-project.org>
>> > Subject: Re: [R] Initializing vector and matrices
>> >
>> > [External Email]
>> >
>> > x <- numeric(0)
>> > for (...) {
>> >     x[length(x)+1] <- ...
>> > }
>> > works.
>> > You can build a matrix by building a vector one element at a time this way, and then reshaping it at the end.  That only works if you don't need it to be a matrix at all times.
>> > Another approach is to build a list of rows.  It's not a matrix, but a list of rows can be a *ragged* matrix with rows of varying length.
>> >
>> > On Wed, 28 Feb 2024 at 21:57, Steven Yen <styen using ntu.edu.tw> wrote:
>> > >
>> > > Is there as way to initialize a vector (matrix) with an unknown length
>> > > (dimension)? NULL does not seem to work. The lines below work with a
>> > > vector of length 4 and a matrix of 4 x 4. What if I do not know
>> > > initially the length/dimension of the vector/matrix?
>> > >
>> > > All I want is to add up (accumulate)  the vector and matrix as I go
>> > > through the loop.
>> > >
>> > > Or, are there other ways to accumulate such vectors and matrices?
>> > >
>> > >  > x<-rep(0,4)  # this works but I like to leave the length open  >
>> > > for (i in 1:3){
>> > > +  x1<-1:4
>> > > +  x<-x+x1
>> > > + }
>> > >  > x
>> > > [1]  3  6  9 12
>> > >
>> > >  > y = 0*matrix(1:16, nrow = 4, ncol = 4); # this works but I like to
>> > > leave the dimension open
>> > >       [,1] [,2] [,3] [,4]
>> > > [1,]    0    0    0    0
>> > > [2,]    0    0    0    0
>> > > [3,]    0    0    0    0
>> > > [4,]    0    0    0    0
>> > >  > for (i in 1:3){
>> > > +   y1<-matrix(17:32, nrow = 4, ncol = 4)
>> > > +   y<-y+y1
>> > > + }
>> > >  > y
>> > >       [,1] [,2] [,3] [,4]
>> > > [1,]   51   63   75   87
>> > > [2,]   54   66   78   90
>> > > [3,]   57   69   81   93
>> > > [4,]   60   72   84   96
>> > >  >
>> > >
>> > > ______________________________________________
>> > > R-help using r-project.org mailing list -- To UNSUBSCRIBE and more, see
>> > > https://stat/
>> > > .ethz.ch%2Fmailman%2Flistinfo%2Fr-help&data=05%7C02%7Ctebert%40ufl.edu
>> > > %7Cdbccaccf29674b10b17308dc39114d38%7C0d4da0f84a314d76ace60a62331e1b84
>> > > %7C0%7C0%7C638447993707432549%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAw
>> > > MDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C0%7C%7C%7C&sdata=
>> > > PtWjcDOnwO7PArVOSdgYbpz8ksjDPK%2Bn9ySyhwQC0gE%3D&reserved=0
>> > > PLEASE do read the posting guide
>> > > http://www.r/
>> > > -project.org%2Fposting-guide.html&data=05%7C02%7Ctebert%40ufl.edu%7Cdb
>> > > ccaccf29674b10b17308dc39114d38%7C0d4da0f84a314d76ace60a62331e1b84%7C0%
>> > > 7C0%7C638447993707438911%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiL
>> > > CJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C0%7C%7C%7C&sdata=Igb16
>> > > CBYgG21HLEDH4I4gfjjFBa3KjDFK8yEZUmBo8s%3D&reserved=0
>> > > and provide commented, minimal, self-contained, reproducible code.
>> >
>> > ______________________________________________
>> > R-help using r-project.org mailing list -- To UNSUBSCRIBE and more, see
>> > 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.
>>
>> ______________________________________________
>> R-help using r-project.org mailing list -- To UNSUBSCRIBE and more, see
>> 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.
>
>______________________________________________
>R-help using r-project.org mailing list -- To UNSUBSCRIBE and more, see
>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.

-- 
Sent from my phone. Please excuse my brevity.

More information about the R-help mailing list