similar to: Discrepancy: R sum() VS C or Fortran sum

My simple functions were to compare the result with the gfortran compiler sum() function. I thought that the Fortran sum could not be less precise than R. I was wrong. I am impressed. The R sum does in fact match the result if we use the Kahan algorithm. P. I am glad to see that R sum() is more accurate than the gfortran compiler sum. On 16/03/18 11:37 AM, luke-tierney at uiowa.edu wrote:

Install the gmp package, run your code, and then try this: bu <- gmp::as.bigq(u) bs4 <- bu[1] + bu[2] + bu[3] + bu[4] + bu[5] s4 <- as.double(bs4) s1 - s4 ## [1] 0 s2[[2]] - s4 ## [1] 7.105427e-15 s3 - s4 ## [1] 7.105427e-15 identical(s1, s4) ## [1] TRUE `bs4` is the exact sum of the binary rationals in your `u` vector; `s4` is the closest double precision to this exact sum.

[[this bounced first, because it has 'help' in the Subject line ... -- Martin Maechler ]] I am a complete novice R programmer. (Though I know C quite well) I am trying to write some R code to do the following simulation. There is a 2-frame "movie" of noise and signal dots. the noise dots have random positions in each frame. The signal dots are placed randomly in frame 1,

I am used to java (well, i dont remember it really well, but anyway) I have having a really difficult time making simple loops to work. I got the following to work: ## ##Creates objects Ux1, Ux2, Ux2 etc. that all contain n numbers in a random distribution ## m<-c(m1,m2,m3,m4,m5,m6,m7,m8,m9,m10)#these are defined as numbers (means) v<-c(v1,v2,v3,v4,v5,v6,v7,v8,v9,v10)#these

rendercheck -f a8r8g8b8 passes as much as on a GK208, and xv appears to work. Very lightly tested. Instead of sticking coordinates into pushbufs, the vertex shader is modified to read them from a constbuf, indexed by vertex id. This approach could be used for all nvc0 generations, but I didn't want to rock the boat. Signed-off-by: Ilia Mirkin <imirkin at alum.mit.edu> --- Note: this

Hi, it's my first time to write a loop with R for my homework. This loop is part of the function. I wanna assign values for hll according to panel [ii,1]=pp. I didn't get any error message in this part. but then when I further calculate another stuff with hll, the function can't return. I think it must be some problem in my loop. Probably something stupid or easy. But I tried to look

Dear all Please see the following module, In module SX3 SX4 and SX5 have similar class Tasklist, but inherit from different class. I will use the following code to connect to different data sources RemoteSX3Model.establish_connection sx3_hash RemoteSX4Model.establish_connection sx4_hash RemoteSX5Model.establish_connection sx5_hash How can I refactor my code in module to look simpler? Thank you

* Kill the A8+A8 hack. Recent enough X servers (>=1.7) fall back to ARGB glyphs for drivers not supporting A8 render targets. * Kill all the global state. It doesn't matter a lot yet but it might if we get multicard working at some point. * Other random clean-ups with no functional changes. Some numbers from x11perf -aa10text -aa24text -comppixwin10 -comppixwin500: * Before, with A

I believe I've addressed all the feedback from the first time around, and also made fixes necessary for GM20x based on testing results. I believe now it should actually work for all GM10x and GM20x. Further, GP10x should be very easy to add, but without someone to actually test I didn't want to claim support for it. Ilia Mirkin (7): exa: add GM10x acceleration support hwdefs: update

This has two purposes : - cleaner code - reduce the diff between nv30 and nv40 exa for a possible nvfx_exa merge ? The main differences seem to be that nv30 uses rect texture format (and does not support repeat on that). Then there are some minor changes in TX_FORMAT RT_FORMAT and TEX_FILTER usage. And NVAccelInitNVx0TCL look complete different. Tested with: ./rendercheck -t

rendercheck -f a8r8g8b8 passes as much as on a GK208, and xv appears to work. Very lightly tested. Instead of sticking coordinates into pushbufs, the vertex shader is modified to read them from a constbuf, indexed by vertex id. This approach could be used for all nvc0 generations, but I didn't want to rock the boat. Signed-off-by: Ilia Mirkin <imirkin at alum.mit.edu> ---

Looks reasonable, some minor comments below. On 10/16/2016 02:06 AM, Ilia Mirkin wrote: > rendercheck -f a8r8g8b8 passes as much as on a GK208, and xv appears to > work. Very lightly tested. > > Instead of sticking coordinates into pushbufs, the vertex shader is > modified to read them from a constbuf, indexed by vertex id. This > approach could be used for all nvc0 generations,

Hi all Forgive me, I'm a total R newbie, and this seems to be a straightforward simple bootstrap problem, but after a whole day of trying to figure out how to do it I'm ready to give up. Part of the problem is that every example and every help page seems to be about doing something more far more complex. I'm got a table with 40 columns and 750 rows. I sum all the values across the

Hi, i am using on my machine Postfix with Cyrus_SASL in smtpd.conf i have saslauthd as method used,but even if i have it , this is what postfix writes into log messages Dec 13 00:02:03 sx1 postfix/smtpd[17833]: auxpropfunc error invalid parameter supplied Dec 13 00:08:07 sx1 postfix/smtpd[17856]: auxpropfunc error invalid parameter supplied Dec 13 00:08:07 sx1 postfix/smtpd[17858]: auxpropfunc

Hello guys, I do not know how it is posible : *vi /etc/postfix/main.cf* # INTERNET HOST AND DOMAIN NAMES # # The myhostname parameter specifies the internet hostname of this # mail system. The default is to use the fully-qualified domain name # from gethostname(). $myhostname is used as a default value for many # other configuration parameters. # #myhostname = host.domain.tld myhostname =

On a PPC machine running Mac OS X, I noticed that creation time is not preserved for files that are transferred (i.e. versus files that are not transferred, or files that have only attribute changes). I discovered two things that work together to cause this: 1) cmp_time (util.c) doesn't check for negative values provided in file1 or file2. If file2 is positive, and file1 is

Hello , i want all my log messages from python to have in /var/log/python Now they are in /var/log/messages: Jun 9 12:10:51 sx1 kernel: klogd 1.4.1, log source = /proc/kmsg started. Jun 9 12:11:01 sx1 python: MM_MEMBERS [it] Nothing to do. Jun 9 12:11:02 sx1 python: MM_MEMBERS [research] Nothing to do. Jun 9 12:12:01 sx1 python: MM_MEMBERS [it] Nothing to do. Jun 9 12:12:01 sx1 python:

I am completely new to Linux and Wine. I've tried to make Wine work since some 1999, but never had patience with it. Tried reading guides, manuals and so on, but things did not work. All Linux versions I've tried have never run any windows-based programs after I installed Wine (and used to install a propper version of wine - that corresponds to Linux version) - what I did is what is said

Hi All, I've been successfully using the with function for analyses and the transform function for multiple transformations. Then I thought, why not use "with" for both? I ran into problems & couldn't figure them out from help files or books. So I created a simplified version of what I'm doing: rm( list=ls() ) x1<-c(1,3,3) x2<-c(3,2,1) x3<-c(2,5,2)

Hi! I will calculate sum??s in the following way: E.g.: a <- rpois(100,20) b <- rpois(100,5) x <- data.frame(cbind(a,b)) # the sum??s should be calculated based on a. attach(x) sort.nace <- unique(sort(x[,1])) sum1 <- matrix(ncol=1, nrow=length(sort.nace)) # I calculate the sum of all values of b, which have the same category in a. Eg.: sum1[1,] <- sum(subset(x,