RE: is.na(v)-b (was: Re: [R] Beginner's query - segmentation fault)
I think the thread ended up with several people (not only me)
feeling certain they didn't like `is.na-` but with the
developers defending it and me not really understanding
why.
Uwe Ligges was going to come up with an example of
`- NA` going wrong (sorry Brian R, I mean behaving
unexpectedly), but never did, and I think the problem
has been fixed. It was apparently a problem with assigning
NAs to an existing factor, but the code for `[-.factor`
looks pretty robust to me [not that I'm at all qualified to say
that, be warned]. Interestingly, at some point both methods
for `is.na-` perform this operation: x[value] - NA. Ahem.
By the way, `is.na(x) - FALSE` will leave x unchanged (including
leaving it as NA ! how bad is that ?!)
-Original Message-
From: Paul Lemmens [mailto:[EMAIL PROTECTED]
Sent: 14 October 2003 16:10
To: [EMAIL PROTECTED]
Subject: RE: is.na(v)-b (was: Re: [R] Beginner's query - segmentation
fault)
Security Warning:
If you are not sure an attachment is safe to open please contact
Andy on x234. There are 0 attachments with this message.
By accident I'm also toying around with NA's, so I started
reading up on
this thread but failed to find a 'concluding' remark or advice. As a
naive
R user I would have loved to see a comment do it like this.
The prevailing opinion seemed to be that is.na() might be
better (safer)
but x - NA is much clearer to understand. Can I relatively safely use
the
easy form, or is it better to remember (the hard way) the
safer version?
Has the discussion continued privately or just stopped here?
Personally I still find the fragments below (taken from the
thread) very
counter intuitive, not to say scary.
x - 1:10
is.na(x) - 1:5
and
is.na(x) - FALSE
It's very hard to understand what happens (as layman) because the
assignment seems to reverse in meaning in the first example (actually
taking indices 1:5 of x and assigning those the value NA)
whereas in the
second case it's not obvious what happens to x: will it get the value
FALSE
or will the original value remain(*).
IMHO the - NA construct is much easier to understand and
should be made
safe in all possible situations (whatever the underlying
safety problem
or
other difficulties might be).
kind regards,
Paul
(*) Such a remark will probably lead to some kind of reprimand because
it's
probably somewhere within the 10e6 manual pages but I'm trying my luck
here.
--
Paul Lemmens
NICI, University of Nijmegen ASCII Ribbon Campaign /\
Montessorilaan 3 (B.01.03)Against HTML Mail \ /
NL-6525 HR Nijmegen X
The Netherlands / \
Phonenumber+31-24-3612648
Fax+31-24-3616066
__
[EMAIL PROTECTED] mailing list
https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Simon Fear
Senior Statistician
Syne qua non Ltd
Tel: +44 (0) 1379 69
Fax: +44 (0) 1379 65
email: [EMAIL PROTECTED]
web: http://www.synequanon.com
Number of attachments included with this message: 0
This message (and any associated files) is confidential and\...{{dropped}}
__
[EMAIL PROTECTED] mailing list
https://www.stat.math.ethz.ch/mailman/listinfo/r-help
RE: is.na(v)-b (was: Re: [R] Beginner's query - segmentation fault)
Hello Simon, --On woensdag 15 oktober 2003 10:08 +0100 Simon Fear [EMAIL PROTECTED] wrote: By the way, `is.na(x) - FALSE` will leave x unchanged (including leaving it as NA ! how bad is that ?!) Twilight Zone (Golden Earring). But with that remark I'm getting off topic, so thank you for your summary. I've already memorized the is.na() construct, so I should be safe for the time being : kind regards, Paul -- Paul Lemmens NICI, University of Nijmegen ASCII Ribbon Campaign /\ Montessorilaan 3 (B.01.03)Against HTML Mail \ / NL-6525 HR Nijmegen X The Netherlands / \ Phonenumber+31-24-3612648 Fax+31-24-3616066 __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
RE: is.na(v)-b (was: Re: [R] Beginner's query - segmentation fault)
By accident I'm also toying around with NA's, so I started reading up on this thread but failed to find a 'concluding' remark or advice. As a naive R user I would have loved to see a comment do it like this. The prevailing opinion seemed to be that is.na() might be better (safer) but x - NA is much clearer to understand. Can I relatively safely use the easy form, or is it better to remember (the hard way) the safer version? Has the discussion continued privately or just stopped here? Personally I still find the fragments below (taken from the thread) very counter intuitive, not to say scary. x - 1:10 is.na(x) - 1:5 and is.na(x) - FALSE It's very hard to understand what happens (as layman) because the assignment seems to reverse in meaning in the first example (actually taking indices 1:5 of x and assigning those the value NA) whereas in the second case it's not obvious what happens to x: will it get the value FALSE or will the original value remain(*). IMHO the - NA construct is much easier to understand and should be made safe in all possible situations (whatever the underlying safety problem or other difficulties might be). kind regards, Paul (*) Such a remark will probably lead to some kind of reprimand because it's probably somewhere within the 10e6 manual pages but I'm trying my luck here. -- Paul Lemmens NICI, University of Nijmegen ASCII Ribbon Campaign /\ Montessorilaan 3 (B.01.03)Against HTML Mail \ / NL-6525 HR Nijmegen X The Netherlands / \ Phonenumber+31-24-3612648 Fax+31-24-3616066 __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
RE: is.na(v)-b (was: Re: [R] Beginner's query - segmentation fault)
-Original Message-
From: Richard A. O'Keefe [mailto:[EMAIL PROTECTED]
snip
The very existence of an is.na- which accepts a logical
vector containing FALSE as well as TRUE ...
And don't forget this is not the only usage of is.na-. In fact it is
designed to take any valid indexing value. For example:
a-1:10
is.na(a) - 1:5
a
[1] NA NA NA NA NA 6 7 8 9 10
Wow. I really hate that. Someone tell me again why this is
better than a[1:5] - NA ??
Simon Fear
Senior Statistician
Syne qua non Ltd
Tel: +44 (0) 1379 69
Fax: +44 (0) 1379 65
email: [EMAIL PROTECTED]
web: http://www.synequanon.com
Number of attachments included with this message: 0
This message (and any associated files) is confidential and\...{{dropped}}
__
[EMAIL PROTECTED] mailing list
https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: is.na(v)-b (was: Re: [R] Beginner's query - segmentation fault)
Richard A. O'Keefe wrote: I am puzzled by the advice to use is.na(x) - TRUE instead of x - NA. ?NA says Function `is.na-' may provide a safer way to set missingness. It behaves differently for factors, for example. However, MAY provide is a bit scary, and it doesn't say WHAT the difference in behaviour is. I must say that is.na(x) - ... is rather repugnant, because it doesn't work. What do I mean? Well, as the designers of SETL who many years ago coined the term sinister function call to talk about f(...)-..., pointed out, if you do f(x) - y then afterwards you expect f(x) == y to be true. So let's try it: x - c(1,NA,3) is.na(x) - c(FALSE,FALSE,TRUE) x [1] 1 NA NA is.na(x) [1] FALSE TRUE TRUE v So I _assigned_ c(FALSE,FALSE,TRUE) to is.na(x), but I _got_ c(FALSE,TRUE, TRUE) instead. ^ That is not how a well behaved sinister function call should work, and it's enough to scare someone off is.na()- forever. The obvious way to set elements of a variable to missing is ... - NA. Wouldn't it be better if that just plain worked? Can someone give an example of is.na()- and -NA working differently with a factor? I just tried it: x - factor(c(3,1,4,1,5,9)) y - x is.na(x) - x==1 y[y==1] - NA x [1] 3NA 4NA 59 Levels: 1 3 4 5 9 y [1] 3NA 4NA 59 Levels: 1 3 4 5 9 Both approaches seem to have given the same answer. What did I miss? As mentioned in another mail to R-help. I'm pretty sure there was (is?) a problem with character (and/or factor) and assignment of NAs, but I cannot (re)produce an example. I think something for the x - NA case has been fixed during the last year. What prevents me to think I'm completely confused is that the is.na()- usage is proposed in: ?NA, S Programming, the R Language Definition manual, R's News file, but I cannot find it in the green book right now. Uwe Ligges __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
RE: is.na(v)-b (was: Re: [R] Beginner's query - segmentation fault)
Note this behaviour:
a-a
a-NA
mode(a)
[1] logical
a-a
is.na(a) - T
mode(a)
[1] character
However after either way of assigning NA to a, is.na(a) is true,
and it prints as NA, so I can't see it's ever likely to matter. [Why
do I say these things? Expect usual flood of examples where it
does matter.]
Also if a is a character vector, a[2] - NA coerces the NA to
as.character(NA); again, just as one would hope/expect.
I have to echo Richard O'K's remark: if - NA can ever go wrong,
is that not a bug rather than a feature?
Simon Fear
Senior Statistician
Syne qua non Ltd
Tel: +44 (0) 1379 69
Fax: +44 (0) 1379 65
email: [EMAIL PROTECTED]
web: http://www.synequanon.com
Number of attachments included with this message: 0
This message (and any associated files) is confidential and\...{{dropped}}
__
[EMAIL PROTECTED] mailing list
https://www.stat.math.ethz.ch/mailman/listinfo/r-help
RE: is.na(v)-b (was: Re: [R] Beginner's query - segmentation fault)
On Wed, 8 Oct 2003, Simon Fear wrote: Note this behaviour: a-a a-NA mode(a) [1] logical a-a is.na(a) - T mode(a) [1] character However after either way of assigning NA to a, is.na(a) is true, and it prints as NA, so I can't see it's ever likely to matter. [Why do I say these things? Expect usual flood of examples where it does matter.] Also if a is a character vector, a[2] - NA coerces the NA to as.character(NA); again, just as one would hope/expect. I have to echo Richard O'K's remark: if - NA can ever go wrong, is that not a bug rather than a feature? I don't think it can ever `go wrong', but it can do things other than the user intends. The intention of is.na- is clearer, and so perhaps user error is less likely? That is the thinking behind the function, anyway. -- Brian D. Ripley, [EMAIL PROTECTED] Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272866 (PA) Oxford OX1 3TG, UKFax: +44 1865 272595 __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
RE: is.na(v)-b (was: Re: [R] Beginner's query - segmentation fault)
Well, that's a convincing argument, but maybe
it's the name that's worrying some of us. Maybe it would be
more intuitive if called set.na (sorry, I mean setNA).
Also is.na- cannot be used to create a new variable of
NAs, so is not a universal method, which is a shame for its
advocates.
I note also that for a vector you can assign a new NA using
either TRUE or FALSE:
a - 1:3
is.na(a[4])-F
a
[1] 1 2 3 NA
For a list, assigning F leaves the new element set to NULL.
Mind you, I suspect this would be a particularly stupid thing
to do, so I'm not going to lose any sleep over R's reaction to it.
-Original Message-
From: Prof Brian Ripley [mailto:[EMAIL PROTECTED]
I don't think it can ever `go wrong', but it can do things other than
the
user intends. The intention of is.na- is clearer, and so
perhaps user
error is less likely? That is the thinking behind the
function, anyway.
Simon Fear
Senior Statistician
Syne qua non Ltd
Tel: +44 (0) 1379 69
Fax: +44 (0) 1379 65
email: [EMAIL PROTECTED]
web: http://www.synequanon.com
Number of attachments included with this message: 0
This message (and any associated files) is confidential and\...{{dropped}}
__
[EMAIL PROTECTED] mailing list
https://www.stat.math.ethz.ch/mailman/listinfo/r-help
RE: is.na(v)-b (was: Re: [R] Beginner's query - segmentation fault)
Also, presumably is.na- could be redefined by the user for particular classes so if you got in the habit of setting NAs that way it would generalize better. --- Date: Wed, 8 Oct 2003 11:49:29 +0100 (BST) From: Prof Brian Ripley [EMAIL PROTECTED] I don't think it can ever `go wrong', but it can do things other than the user intends. The intention of is.na- is clearer, and so perhaps user error is less likely? That is the thinking behind the function, anyway. ___ No banners. No pop-ups. No kidding. Introducing My Way - http://www.myway.com __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
RE: is.na(v)-b (was: Re: [R] Beginner's query - segmentation fault)
Simon Fear [EMAIL PROTECTED] suggested that a-a a-NA mode(a) [1] logical a-a is.na(a) - T mode(a) [1] character might be a relevant difference between assigning NA and using is.na. But the analogy is flawed: is.na(x) - operates on the _elements_ of x, while x - affects the variable x. When you assign NA to _elements_ of a vector, the mode does not change: a - a is.na(a) - TRUE mode(a) [1] character b - b b[TRUE] - NA mode(b) [1] character c - c c[1] - NA mode(c) [1] character __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
RE: is.na(v)-b (was: Re: [R] Beginner's query - segmentation fault)
Concerning x[i] - NA vs is.na(x[i]) - TRUE
Brian Ripley wrote:
I don't think it can ever `go wrong', but it can do things other
than the user intends.
If the user writes x[i] - NA, the user has clearly indicated his intention
that the i element(s) of x should become NA. There isn't any clearer way to
say that. The only way it could ever do something other than the user
intends is if the mode of x changes or the selected elements are set to
something other than NA.
The ?NA help page *hints* that this might be the case, but does not give
an example.
The question remains, *WHAT* can x[i]-NA do that might be other than
what the user intends? An example (especially one added to the ?NA help)
would be very useful.
The intention of is.na- is clearer,
I find this extremely puzzling. x[i] - NA is an extremely clear and
obvious way of saying I want the i element(s) of x to become NA.
is.na(x) - ... is not only an indirect way of doing this, it is a way
which is confusing and error-prone.
Bear in mind that one way of implementing something is is.na() would be
to associate a bit with each element of a vector; is.na() would test and
is.na-() would set that bit. It would be possible to have a language
exactly like R -except- that
x - 1
is.na(x) - TRUE
x
= NA
is.na(x) - FALSE
x
= 1
would work. The very existence of an is.na- which accepts a logical
vector containing FALSE as well as TRUE strongly suggests this. But it
doesn't work like that. As I've pointed out,
is.logical(m) length(m) == length(x) done{is.na(x) - m}
= identical(is.na(x), m)
is the kind of behaviour that has been associated with well-behaved
sinister function calls for several decades, and yet this is not a fact
about R.
and so perhaps user error is less likely?
I see no reason to believe this; the bad behaviour of is.na- surely
makes user error *more* likely rather than *less* likely.
__
[EMAIL PROTECTED] mailing list
https://www.stat.math.ethz.ch/mailman/listinfo/r-help
Re: is.na(v)-b (was: Re: [R] Beginner's query - segmentation fault)
Tongue in cheek But surely is.na(x) - is.na(x) is clearer than x[is.na(x)] - NA (neither of which is a no-op). /Tongue in cheek __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
is.na(v)-b (was: Re: [R] Beginner's query - segmentation fault)
I am puzzled by the advice to use is.na(x) - TRUE instead of x - NA. ?NA says Function `is.na-' may provide a safer way to set missingness. It behaves differently for factors, for example. However, MAY provide is a bit scary, and it doesn't say WHAT the difference in behaviour is. I must say that is.na(x) - ... is rather repugnant, because it doesn't work. What do I mean? Well, as the designers of SETL who many years ago coined the term sinister function call to talk about f(...)-..., pointed out, if you do f(x) - y then afterwards you expect f(x) == y to be true. So let's try it: x - c(1,NA,3) is.na(x) - c(FALSE,FALSE,TRUE) x [1] 1 NA NA is.na(x) [1] FALSE TRUE TRUE v So I _assigned_ c(FALSE,FALSE,TRUE) to is.na(x), but I _got_ c(FALSE,TRUE, TRUE) instead. ^ That is not how a well behaved sinister function call should work, and it's enough to scare someone off is.na()- forever. The obvious way to set elements of a variable to missing is ... - NA. Wouldn't it be better if that just plain worked? Can someone give an example of is.na()- and -NA working differently with a factor? I just tried it: x - factor(c(3,1,4,1,5,9)) y - x is.na(x) - x==1 y[y==1] - NA x [1] 3NA 4NA 59 Levels: 1 3 4 5 9 y [1] 3NA 4NA 59 Levels: 1 3 4 5 9 Both approaches seem to have given the same answer. What did I miss? __ [EMAIL PROTECTED] mailing list https://www.stat.math.ethz.ch/mailman/listinfo/r-help
