I had the same FIRST impression which could have made sense as in writing
the data back out to a file with a similar name and that would have been
trivial. I mean if you had exactly three, a loop would not even be
particularly useful versus writing two lines, copying them and editing the
latter two sets.
But the OP could have done a better job explaining what they wanted it all
FOR. Just writing out the same file contents would be nonsensical unless
some change was made such as writing in a different format. But I noted the
output names were not something like NAME.csv and although the code looked
wrong, it soon became clear they wanted to create not files but clearly a
data.frame as that is produced by read.csv. The language usage was simply
sloppy.
The names chosen made it easier as all you needed to do was replace
"2010midata" with "bop" and easy enough to do in a loop or several other
ways but what the OP missed was how to do something that is doable in an
interpreted language like R but possibly not in many compiled languages.
So, as others pointed out, sometimes the person doing this has become wedded
to ONE WAY of doing something rather than considering if some other way is a
better choice. In this case, a consensus is that a data structure such as a
list of data.frames would be a better choice and even that it could be done
without an explicit loop such as this version where you can set N to any
value, and it will get files with a suffix ranging from 1 to N, albeit this
does not allow you to find something like 007 without modification:
N <- 3
list_of_df <- lapply(paste0("2010midata", 1:N, ".csv"), read.csv)
An individual data.frame can be accessed using [[]] notation as in:
> class(list_of_df[[1]])
[1] "data.frame"
Of course if you have additional values to pass in to read.csv, add them in
the ... after the first args and so on.
The next part, badly written as:
paste0("bop",im)<-boprobit(eqs,mydata,wt=weight,method="BHHH",tol=0,reltol=0
,gradtol=1e-5,Fisher=TRUE)
Not can use a similar construction in which the 1:N is replaced by
list_of_df and the function by boprobit and then the remaining arguments in
the ...
But that is not the way many initially think. Let alone thinking in nested
fashion as in:
Results <- lapply(X=lapply(X=paste0("2010midata", 1:N, ".csv"),
FUN=read.csv), FUN=boprobit,
,wt=weight,method="BHHH",tol=0,reltol=0,gradtol=1e-5,Fisher=TRUE))
Unfortunately, the latter is WRONG as boprobit seems to take a second
argument of the data.frame with whatever "eqs" is as the first argument.
This slightly nonstandard version would require some gimmick such as an
accessory function that rearranges what it gets as arguments and then calls
boprobit.
But, as noted, this is one of many examples where the user comes up with a
"solution" that is fairly rarely needed even if it can be done, rather than
one that is a bit more abstract and yet powerful.
Yet, as noted, R can do things we often have no need for and this is an
example. Never mind how it can be a tad dangerous to assign to extra
variable names that might step on existing names as compared to creating
anonymous data structures with few or no names as everything is largely
positional.
I will end with a slight variant. Imagine making a named list, perhaps
extended in each part of the loop used, so that you it looks like
Mylist <- list(name1=value, name2=value, name3=value)
This could then be addressed using Mylist[[1]] or Mylist$name1
interchangeably. But as all names would be internal to the list, it is
name-safe.
-----Original Message-----
From: R-help <r-help-boun...@r-project.org> On Behalf Of Richard O'Keefe
Sent: Monday, June 24, 2024 9:05 PM
To: Steven Yen <st...@ntu.edu.tw>
Cc: R-help Mailing List <r-help@r-project.org>; Steven Yen
<sye...@gmail.com>
Subject: Re: [R] Naming output file
The subject line says (capitalisation changed) "name output FILE"
but I see no attempt in the sample code to create an output FILE.
I was expecting to see something like
write(<whatever>, file = paste0("bop",im))
On Mon, 24 Jun 2024 at 23:41, Steven Yen <st...@ntu.edu.tw> wrote:
>
> I would like a loop to
>
> (1) read data files 2010midata1,2010midata2,2010midata3; and
>
> (2) name OUTPUT bop1,bop2,bop3.
>
> I succeeded in line 3 of the code below,
>
> BUT not line 4. The error message says:
>
> Error in paste0("bop", im) <- boprobit(eqs, mydata, wt = weight, method
> = "NR", : target of assignment expands to non-language object Please
> help. Thanks.
>
> m<-3
> for (im in 1:m) {
> mydata<-read.csv(paste0("2010midata",im,".csv"))
>
paste0("bop",im)<-boprobit(eqs,mydata,wt=weight,method="BHHH",tol=0,reltol=0
,gradtol=1e-5,Fisher=TRUE)
> }
>
>
>
> [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help@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@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@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.
