Yes, that is what I meant. I will commit that.
On Wed, 29 Mar 2006, Andrew Robinson wrote:
> I tried it. That is, in qhyper.c I replaced
>
> p *= 1 - 64*DBL_EPSILON;
>
> with
>
> p *= 1 - 1000*DBL_EPSILON;
>
> and recompiled. 2.3.0 alpha now passses make check-all and returns
> sensible result
This message is in MIME format. The first part should be readable text,
while the remaining parts are likely unreadable without MIME-aware tools.
--27464147-1125417764-1143612462=:26272
Content-Type: TEXT/PLAIN; charset=iso-8859-1; format=flowed
Content-Transfer-Encoding: 8BIT
Please read th
Hi, I am just trying it now. Suppose I copy the first few lines of
the table at:
http://www.pricelesswarehome.org/2006/CumulativePL.php
into the clipboard by selecting them with the mouse in IE and
pressing ctrl-C. Now I can just go to Excel, click on a cell
and press ctrl-V and they get pasted
Peter Dalgaard biostat.ku.dk> writes:
>
> I'm not even sure we want to fix this up. I recall some nasty issues
> with DF that have no proper solution that way - an observation with a
> tiny weight represents an observation with a large variance and
> contributes 1DF to the residual, with weight
> -Original Message-
> From: Simon Urbanek [mailto:[EMAIL PROTECTED]
> Sent: Friday, March 24, 2006 8:04 AM
> To: Matthew Beason
> Cc: r-devel@r-project.org
> Subject: Re: [Rd] R make install and demo(graphics) issue
>
> Matthew,
>
> On Mar 20, 2006, at 1:52 PM, Matthew Beason wrote:
>
I tried it. That is, in qhyper.c I replaced
p *= 1 - 64*DBL_EPSILON;
with
p *= 1 - 1000*DBL_EPSILON;
and recompiled. 2.3.0 alpha now passses make check-all and returns
sensible results on both of my FreeBSD 6.1 machines.
> Rhyper <- 1:20
> Phyper <- phyper (Rhyper, m = 40, n = 30,
Certainly. I assume that you mean qhyper.c rather than qgamma.c. Can
you please be more explicit as to how I can try your suggestion. Do
you mean that I should try:
p *= 1 - 1000*DBL_EPSILON;
?
Thanks.
On Tue, Mar 28, 2006 at 01:53:09PM +0100, Prof Brian Ripley wrote:
> qgamma.c comtains the
Full_Name: Jason Trimble
Version: 2.2.1
OS: Windows XP and RedHat EL4
Submission from: (NULL) (209.166.128.18)
Whenever a MySQL query returns a Decimal result to R I get the following warning
in R:
Warning message:RS-DBI driver warning: (unrecognized MySQL field type 246 in
column 1)
I get thi
[EMAIL PROTECTED] writes:
> First, R-bugs is not for asking questions, only for reporting things you
> are *certain* are bugs: see the R FAQ.
>
> Why are you using weights to omit cases? If you had used subset, this
> would have worked. The problem is the use of zero weights, which are not
>
First, R-bugs is not for asking questions, only for reporting things you
are *certain* are bugs: see the R FAQ.
Why are you using weights to omit cases? If you had used subset, this
would have worked. The problem is the use of zero weights, which are not
intended to be used in this way. We c
BTW, sections in .texi manuals do not have numbers, so please use the
node name.
You can use .First.lib.
On Tue, 28 Mar 2006, Henrik Bengtsson wrote:
> Hi,
>
> how do I set options() when loading a package *without* a name space?
> Is it possible?
>
> I though this one was a common question, b
On 3/28/2006 10:18 AM, Henrik Bengtsson wrote:
> Hi,
>
> how do I set options() when loading a package *without* a name space?
> Is it possible?
Use .First.lib(). It is mentioned in the R-exts section you looked in
below, but its purpose is not.
Duncan Murdoch
>
> I though this one was a co
Don't worry. Sorry for that, I should of course know that it is in
.First.lib() as documented Section 1.1.4 on "Package subdirectories"
in "Writing R Extensions"; "A common use is to call library.dynam()
inside .First.lib() to load compiled code: another use is to call
those functions with side ef
Full_Name: Robert Pusz
Version: 2.2.1
OS: Windows
Submission from: (NULL) (157.25.9.126)
Hello,
In my opinion something is wrong with 'weights' option in glm.
My code is following:
###begin of the code
cl <- c(5012, 106, 3410, 5655, 1092, 1513, 557, 1351, 3133, 2063, 3257,
4179, 5582, 5900,
Hi,
how do I set options() when loading a package *without* a name space?
Is it possible?
I though this one was a common question, but I could not find it in
the FAQ, in the help nor in the r-help/r-devel archives. Section
1.6.3 on "Load hooks" in "Writing R Extensions" says that this should
be
qgamma.c comtains the line
p *= 1 - 64*DBL_EPSILON;
which seems to be too small nowadays. I find for example
> p <- phyper(7, m = 40, n = 30, k = 20)
>
> qhyper(p*(1+5e-16), m = 40, n = 30, k = 20)
[1] 7
> qhyper(p*(1+6e-16), m = 40, n = 30, k = 20)
[1] 8
so it seems far more sensitive t
Addendum: the following is true for both 2.2.1 and 2.3.0
>
> > qhyper(0.01796062766370490, m = 40, n = 30, k = 20)
> [1] 7
> > qhyper(0.01796062766370491, m = 40, n = 30, k = 20)
> [1] 8
>
--
Andrew Robinson
Department of Mathematics and StatisticsTel: +61-3-8344-9763
Universit
This is from 2.2.1:
> Phyper <- phyper(7, m = 40, n = 30, k = 20)
> print(Phyper, digits=16)
[1] 0.01796062766370490
This is from 2.3.0:
> Phyper <- phyper(7, m = 40, n = 30, k = 20)
> print(Phyper, digits=16)
[1] 0.01796062766370491
> qhyper(Phyper, m = 40, n = 30, k = 20)
[1] 8
Then if I sav
Then we need you to dig in to find out why. I'd start by seeing if Phyper
had changed (either by printing it to 16dp or by saving it from 2.2.1 and
reloading into 2.3.0).
On Tue, 28 Mar 2006, Andrew Robinson wrote:
> I get:
>
>> Rhyper <- scan()
> 1: 16 11 11 15 11 13 13 12 13 10 10 7 11 14 1
RISPOSTA AUTOMATICA
L'indirizzo [EMAIL PROTECTED] non è più attivo.
Se desideri contattarci ti invitiamo a farlo attraverso questo indirizzo:
http://www.chatta.it/contattaci.asp
Cordiali Saluti
Lo Staff di www.Chatta.it
__
R-devel@r-project.org m
Merci, François!
> "FrPi" == François Pinard <[EMAIL PROTECTED]>
> on Tue, 28 Mar 2006 03:00:19 +0200 (CEST) writes:
FrPi> Hi, people. Here is a transcript of a "R --vanilla" session:
FrPi>
==>
FrPi> R
I get:
> Rhyper <- scan()
1: 16 11 11 15 11 13 13 12 13 10 10 7 11 14 13 9 14 13 13 11
21:
Read 20 items
> Phyper <- phyper (Rhyper, m = 40, n = 30, k = 20)
> qhyper(Phyper, m = 40, n = 30, k = 20)
[1] 16 11 11 15 11 13 13 12 13 10 10 8 11 14 13 9 14 1
On Tue, 28 Mar 2006, Andrew Robinson wrote:
> You're welcome. You are correct. d-p-q-r-tests.Rout.fail
> shows:
>
>> All.eq(Rhyper, qhyper (Phyper, m = 40, n = 30, k = 20))
> [1] "Mean scaled difference: 0.0833"
Yes, please run the lines below, e.g.
Rhyper <- scan()
16 11 11 15
You're welcome. You are correct. d-p-q-r-tests.Rout.fail
shows:
> All.eq(Rhyper, qhyper (Phyper, m = 40, n = 30, k = 20))
[1] "Mean scaled difference: 0.0833"
Let me know if/how I can further assist.
Andrew
On Tue, Mar 28, 2006 at 09:03:48AM +0100, Prof Brian Ripley wrote:
> "Andrew" == Andrew Robinson <[EMAIL PROTECTED]>
> on Tue, 28 Mar 2006 17:55:37 +1000 writes:
Andrew> Hi Developers,
Andrew> The alpha, compiles successfully, but it is failing make check-all
(on
Andrew> two seperate machines, both FreeBSD 6.1).
Andrew> Here is the v
Thanks for checking.
Please look in d-p-q-r-tests.Rout.fail and see what immediately preceeds
the line
[1] "Mean scaled difference: 0.0833"
Some experimentation suggests it is
> All.eq(Rhyper, qhyper (Phyper, m = 40, n = 30, k = 20))
If so, we have
Rhyper <- scan()
16 11 11 1
26 matches
Mail list logo
