Hi all,

This message is for the 70 people from the old ANU alliance list who
have not yet signed up for the new meetup.com list. I will not be
sending any more emails to this old list. In fact we will probably try
to delete the old list soon to avoid confusion.

I have just announced a meeting of the Canberra R Users Group for this
coming Monday. You can see the details here:
http://www.meetup.com/Canberra-R-Users-Group/calendar/14175846/


Cheers
-Felix


---------- Forwarded message ----------
From: Felix Andrews <felix-fJbp3clVp5DYtjvyW6yDsg@public.gmane.org>
Date: 22 July 2010 11:44
Subject: [CanberraRUG] New Meetup: MDS in large-scale text corpus visualization
To: CanberraRUG-announce-qgLDrG2BSrPQT0dZR+AlfA@public.gmane.org


Announcing a new Meetup for Canberra R Users Group!

What: MDS in large-scale text corpus visualization

When: Monday, July 26, 2010 3:30 PM

Where: CMIS Seminar Room (room S206)
2nd floor, Computer Science building ANU
Canberra

Glen Newton of Carleton University, Canada is visiting ANU and has
kindly offered to give a talk:

We examine the scalability and validity of semantically mapping
(visualizing) journals in a large scale (5.7+ million) science,
technology and medical article digital library. This work is part of a
larger research effort to evaluate semantic journal and article
mapping for search query results refinement and visual
contextualization in a large scale digital library.

Starting from the 512-space created by the semantic analysis, R's mds
is used to transform the space into 2-D for the visualization. A
visualization of ~2200 journals is presented, and a discussion of
future work involving R.

________________________________

Glen Newton is a graduate student at Carleton University, Ottawa,
Canada. From 1998 - 2010 he was at the National Research Council
Canada, where he was a software group leader, a research scientist,
and a research group leader. His research interests at the NRC and at
Carleton are in the area of knowledge extraction from the research
literature, through text mining, augmented search and visualization.

His interests also include information retrieval, Semantic Web,
digital libraries, scientific data issues,
geographic information system (GIS), and software standards. He is a
committee member of ISO JTC1/SC38 (Distributed computing) and the
Canadian National Committee for CODATA. From 2001 to 2010 he was the
NRC W3C advisory committee representative.

Previous to working at the NRC, Glen worked at the National Atlas of
Canada, Natural Resources Canada, in the area of GIS R&D and is the
author of NAISMap, an early (1994) online mapping tool.

RSVP to this Meetup:
http://www.meetup.com/Canberra-R-Users-Group/calendar/14175846/

This automatic notification message was sent by Alliance (http://alliance.anu.edu.au/portal) from the Canberra R Users site.
You can modify how you receive notifications at My Workspace > Preferences.

Hi all,

We have a wonderful, friendly new website and email list at
http://www.meetup.com/Canberra-R-Users-Group/

Future announcements will happen on meetup.com, and not on this email
list (ANU Alliance). Therefore if you want to remain involved in the
group, you will need to sign up to the new site.

When you sign up you will be asked to briefly answer 2 short questions
about yourself; this helps others to get to know you. Similarly, you
are encouraged (though not required) to upload a photo of yourself.

The use of meetup.com was made possible by sponsorship from
Presciient, thanks to Eugene Dubossarsky (of the Sydney group, who
incidentally held their first meeting yesterday).

All the best,
-Felix (and on behalf of Alec, John and Graham).

--
Felix Andrews
Integrated Catchment Assessment and Management (iCAM) Centre
Fenner School of Environment and Society [Bldg 48a]
The Australian National University
Canberra ACT 0200 Australia
M: +61 410 400 963
T: + 61 2 6125 4670
E: felix.andrews-FCV4sgi5zeUQrrorzV6ljw@public.gmane.org
CRICOS Provider No. 00120C

This automatic notification message was sent by Alliance (http://alliance.anu.edu.au/portal) from the Canberra R Users site.
You can modify how you receive notifications at My Workspace > Preferences.

xmlns:v="urn:schemas-microsoft-com:vml" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:w="urn:schemas-microsoft-com:office:word" xmlns:st1="urn:schemas-microsoft-com:office:smarttags" xmlns="http://www.w3.org/TR/REC-html40">

This automatic notification message was sent by Alliance (http://alliance.anu.edu.au/portal) from the Canberra R Users site.
You can modify how you receive notifications at My Workspace > Preferences.

xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:w="urn:schemas-microsoft-com:office:word" xmlns:st1="urn:schemas-microsoft-com:office:smarttags" xmlns="http://www.w3.org/TR/REC-html40">

Re: [Canberra R Users] Open question/discussion session, July 28, 3pm, CMIS - Please note the time change!

This automatic notification message was sent by Alliance (http://alliance.anu.edu.au/portal) from the Canberra R Users site.
You can modify how you receive notifications at My Workspace > Preferences.

I found I needed to highlight table cells in a Sweave (LaTeX)
document, so I wrote this function to set some table cells in bold
face. Hopefully someone else will find it useful too.

http://gist.github.com/452201

Cheers
-Felix

--
Felix Andrews / 安福立
Integrated Catchment Assessment and Management (iCAM) Centre
Fenner School of Environment and Society [Bldg 48a]
The Australian National University
Canberra ACT 0200 Australia
M: +61 410 400 963
T: + 61 2 6125 4670
E: felix.andrews-FCV4sgi5zeUQrrorzV6ljw@public.gmane.org
CRICOS Provider No. 00120C
-- --
http://www.neurofractal.org/felix/

This automatic notification message was sent by Alliance (http://alliance.anu.edu.au/portal) from the Canberra R Users site.
You can modify how you receive notifications at My Workspace > Preferences.

Hi all,

Remember to come to Monday's R User Group meeting... and for extra
points, think about something to discuss / ask / say.

Cheers
-Felix


On 16 June 2010 09:57, <Alec.Zwart-uXK74I8fpZ0@public.gmane.org> wrote:
> == Canberra R Users Group: http://canrug.togaware.com/ ==
>
> Morning everyone,
>
> For the next CanRUG meeting, we thought we’d have an ‘open’ session where
> people
>
> can bring along their R questions/problems for the group to discuss and
> suggest
>
> solutions or pointers that might help.   Once again, we’ll be holding this
> meeting at
>
> CMIS at ANU (details below).
>
>
>
> We’ll be holding the meeting on the 28th June, but please note that we will
> be beginning
>
>  the session half an hour earlier than usual, at 3pm, and hence finishing at
> 4.30pm.
>
>
> DATE: Monday 28th June
> TIME: 3pm  - Please note the time change!
> LOCATION: CMIS Seminar Room (room S206), CSIT building, ANU
> 2nd floor of the ANU Computer Science & Information Technology Building,
> Bldg 108 http://campusmap.anu.edu.au/displaybldg.asp?no=108
> CSIRO CMIS is on the left (roughly east) side of the building as you
> enter the front door - however, take the stairs to the second floor to
> enter.
>
> SPEAKER: Anyone!
> TITLE: R open question/problem discussion session
>
> We’ll be holding an open session where people can ask questions about R or
> their R
>
> problems for the group to discuss or suggest possible solutions.
>
>
>
> Alexander (Alec) Zwart
> CSIRO Mathematics, Informatics and Statistics
> GPO Box 664
> Canberra ACT 2601, Australia
> Email : Alec.Zwart-uXK74I8fpZ0@public.gmane.org
> Phone : +61 2 62167010
> Fax : +61 2 62167007
>
>
>
>
>
> ________________________________
> This automatic notification message was sent by Alliance
> (http://alliance.anu.edu.au/portal) from the Canberra R Users site.
> You can modify how you receive notifications at My Workspace > Preferences.
>



--
Felix Andrews / 安福立
Integrated Catchment Assessment and Management (iCAM) Centre
Fenner School of Environment and Society [Bldg 48a]
The Australian National University
Canberra ACT 0200 Australia
M: +61 410 400 963
T: + 61 2 6125 4670
E: felix.andrews-FCV4sgi5zeUQrrorzV6ljw@public.gmane.org
CRICOS Provider No. 00120C
-- --
http://www.neurofractal.org/felix/

This automatic notification message was sent by Alliance (http://alliance.anu.edu.au/portal) from the Canberra R Users site.
You can modify how you receive notifications at My Workspace > Preferences.

Eugene and David, thankyou for your generous offers!
(I didn't realise my off-hand remark would result in a deluge of cash...)

I suggest we discuss it at the next meeting (June 28th, 3pm).

Cheers
-Felix


On 17 June 2010 12:54, Eugene Dubossarsky <eugene-QF7pCbK58DS8rjiVs5Nzzw@public.gmane.org> wrote:
>
> Presciient is willing to sponsor monthly meetup.com  costs for both groups for the next 2 years
> Sent from my iPhone
> On Jun 17, 2010, at 12:11 PM, <David.Lovell-uXK74I8fpZ0@public.gmane.org> wrote:
>
> I can find $12/month to help foster a stronger R-user community
>
> DL
>
>
>
> ________________________________
>
> From: foolish.android-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org [mailto:foolish.android-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org] On Behalf Of Felix Andrews
> Sent: Thursday, 17 June 2010 11:03 AM
> To: Canberra R User Group
> Subject: [Canberra R Users] SURF
>
>
>
> See Eugene's message below announcing the Sydney Users of R Forum (SURF) website.
>
> I am jealous: their site is hosted on Meetup, which is soooooo much better than "ANU Alliance" (which is just a joke in terms of usability). If only we could figure out a way to pay the US$12 per month that Meetup charges.
>
> ---------- Forwarded message ----------
> From: Eugene Dubossarsky <eugene-QF7pCbK58DS8rjiVs5Nzzw@public.gmane.org>
> Date: 17 June 2010 10:33
> Subject: CANRUG and SURF
> To: felix.andrews-FCV4sgi5zeUQrrorzV6ljw@public.gmane.org
>
> Dear Felix,
>
> Could you please add me to the CANRUG mailing list. While based in Sydney,  I am frequently in Canberra and most interested in attending CANRUG.
>
> Also,  I would be obliged if you would let your members know about the newly formed Sydney Users of R Forum (SURF)
>
> which they can join through meetup or facebook:
>
> http://www.meetup.com/R-Users-Sydney/
>
> best regards,
>
> Eugene
>
> Dr Eugene Dubossarsky FIAPA
>
> mobile: +61 414 57 33 22
>
> email: eugene-QF7pCbK58DS8rjiVs5Nzzw@public.gmane.org
>
> skype: eugene.dubossarsky
>
> Presciient
>
> Forecasting • Analytics Strategy & Communications • Advanced Analytics • Capability Building • Value from  Analytics • Real Business Intelligence • Accountable Knowledge Management • Decision Performance Management • Data Reconnaissance
>
>
>
>
>
> ________________________________
>
> This automatic notification message was sent by Alliance (http://alliance.anu.edu.au/portal) from the Canberra R Users site.
> You can modify how you receive notifications at My Workspace > Preferences.
>
> ________________________________
> This automatic notification message was sent by Alliance (http://alliance.anu.edu.au/portal) from the Canberra R Users site.
> You can modify how you receive notifications at My Workspace > Preferences.
>
> ________________________________
> This automatic notification message was sent by Alliance (http://alliance.anu.edu.au/portal) from the Canberra R Users site.
> You can modify how you receive notifications at My Workspace > Preferences.



--
Felix Andrews / 安福立
Integrated Catchment Assessment and Management (iCAM) Centre
Fenner School of Environment and Society [Bldg 48a]
The Australian National University
Canberra ACT 0200 Australia
M: +61 410 400 963
T: + 61 2 6125 4670
E: felix.andrews-FCV4sgi5zeUQrrorzV6ljw@public.gmane.org
CRICOS Provider No. 00120C
--
http://www.neurofractal.org/felix/

This automatic notification message was sent by Alliance (http://alliance.anu.edu.au/portal) from the Canberra R Users site.
You can modify how you receive notifications at My Workspace > Preferences.

FYI: a "Sydney Users of R Forum" -- suRf -- has just started up, apparently.



---------- Forwarded message ----------
From: Louise Lor <louise.spisz-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
Date: 15 June 2010 23:49
Subject: Re: R users groups on the R wiki
To: Felix Andrews <felix.andrews-FCV4sgi5zeUQrrorzV6ljw@public.gmane.org>


Thanks for your prompt response, Felix

And wow - amazingly easy to add our group to the list
(we're calling ourselves suRf - Sydney Users of R Forum).

We're very impressed with the CANRUG website and list of talks you've had.
Good to know that there's another more established R users group in Australia.

Cheers,
Louise


On Tue, Jun 15, 2010 at 11:31 PM, Felix Andrews
<felix.andrews-FCV4sgi5zeUQrrorzV6ljw@public.gmane.org> wrote:
>
> Hi Louise,
>
> The nice thing about wikis is that you can edit them yourself
>
> Register at http://rwiki.sciviews.org/doku.php?id=rugs:r_user_groups&do=register
> then you should be able to add the info to the page (though I haven't
> tried it myself).
>
> Otherwise, I think it is Phillipe Grosjean that maintains the site.
>
> It's good to hear about the Sydney user group!
>
> All the best,
> -Felix
>
> On 15 June 2010 23:08, Louise Lor <louise.spisz-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
> > Hi Felix,
> >
> > I saw your name as key contact point for CANRUG.
> >
> > Eugene Dubossarksy & I have just established an R users group in Sydney and
> > are looking to publise it's existence.
> >
> > In particular, we wish to get it listed on the R wiki.
> > http://rwiki.sciviews.org/doku.php?id=rugs:r_user_groups
> >
> > Do you know or anyone at R know the name of the contact person who can do
> > this for us?
> >
> > Cheers & Thanks,
> > Louise Lor
> >
>
>



--
Felix Andrews / 安福立
Integrated Catchment Assessment and Management (iCAM) Centre
Fenner School of Environment and Society [Bldg 48a]
The Australian National University
Canberra ACT 0200 Australia
M: +61 410 400 963
T: + 61 2 6125 4670
E: felix.andrews-FCV4sgi5zeUQrrorzV6ljw@public.gmane.org
CRICOS Provider No. 00120C
-- --
http://www.neurofractal.org/felix/

This automatic notification message was sent by Alliance (http://alliance.anu.edu.au/portal) from the Canberra R Users site.
You can modify how you receive notifications at My Workspace > Preferences.

xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:w="urn:schemas-microsoft-com:office:word" xmlns:st1="urn:schemas-microsoft-com:office:smarttags" xmlns="http://www.w3.org/TR/REC-html40">

[Canberra R Users] Felix Andrews on Lattice graphics, May 17, 3:30, CMIS

This automatic notification message was sent by Alliance (http://alliance.anu.edu.au/portal) from the Canberra R Users site.
You can modify how you receive notifications at My Workspace > Preferences.

(Note that Char is doing a PhD in econometrics at La Trobe Uni, not
just a random person).

You are passing a data frame (with one column) to gpd.fit(), but it
expects a plain numeric vector. Try

pval <- p$V1
or
pval <- p[,1]

Also, I think the gpd.fit() function expects your data to not contain
any missing values. If it is reasonable to remove them (think about it
- it might bias the results), you could do it like this:

pval <- pval[!is.na(pval)]
or
pval <- na.omit(pval)

Otherwise, you might just set them to 0 if you expect that any missing
values are below the threshold. Or select the longest unbroken
sub-sequence using na.contiguous().

Regards
-Felix


On 14 June 2010 14:10, char leung <charleung-PkbjNfxxIARGWvitb5QawA@public.gmane.org> wrote:
> Dear all,
>
> May I get some help about using "ismev" please?
>
> Here's what I have,
>
>> library(ismev)
>> p<-read.table("c:/users/char/desktop/p.txt")
>> str(p)
> 'data.frame':   53855 obs. of  1 variable:
>  $ V1: num  104 90.2 106 105 102 110 103 150 105 110 ...
>> t<-seq(1:53855)
>> set.xy<-data.frame(t,p)
>> plot(set.xy)
>> mrl.plot(p)
>> gpd.fit(p,150)
> Error in if (min(y) <= 0) l <- 10^6 else { :
>   missing value where TRUE/FALSE needed
>
>
> And for the function gpd.fit (I've highlighted the relevant part that may go
> wrong),
>
>> getAnywhere(gpd.fit)
> A single object matching ‘gpd.fit’ was found
> It was found in the following places
>   package:ismev
> with value
>
> function (xdat, threshold, npy = 365, ydat = NULL, sigl = NULL,
>     shl = NULL, siglink = identity, shlink = identity, siginit = NULL,
>     shinit = NULL, show = TRUE, method = "Nelder-Mead", maxit = 10000,
>     ...)
> {
>     z <- list()
>     npsc <- length(sigl) + 1
>     npsh <- length(shl) + 1
>     n <- length(xdat)
>     z$trans <- FALSE
>     if (is.function(threshold))
>         stop("`threshold' cannot be a function")
>     u <- rep(threshold, length.out = n)
>     if (length(unique(u)) > 1)
>         z$trans <- TRUE
>     xdatu <- xdat[xdat > u]
>     xind <- (1:n)[xdat > u]
>     u <- u[xind]
>     in2 <- sqrt(6 * var(xdat))/pi
>     in1 <- mean(xdat, na.rm = TRUE) - 0.57722 * in2
>     if (is.null(sigl)) {
>         sigmat <- as.matrix(rep(1, length(xdatu)))
>         if (is.null(siginit))
>             siginit <- in2
>     }
>     else {
>         z$trans <- TRUE
>         sigmat <- cbind(rep(1, length(xdatu)), ydat[xind, sigl])
>         if (is.null(siginit))
>             siginit <- c(in2, rep(0, length(sigl)))
>     }
>     if (is.null(shl)) {
>         shmat <- as.matrix(rep(1, length(xdatu)))
>         if (is.null(shinit))
>             shinit <- 0.1
>     }
>     else {
>         z$trans <- TRUE
>         shmat <- cbind(rep(1, length(xdatu)), ydat[xind, shl])
>         if (is.null(shinit))
>             shinit <- c(0.1, rep(0, length(shl)))
>     }
>     init <- c(siginit, shinit)
>     z$model <- list(sigl, shl)
>     z$link <- deparse(substitute(c(siglink, shlink)))
>     z$threshold <- threshold
>     z$nexc <- length(xdatu)
>     z$data <- xdatu
>     gpd.lik <- function(a) {
>         sc <- siglink(sigmat %*% (a[seq(1, length = npsc)]))
>         xi <- shlink(shmat %*% (a[seq(npsc + 1, length = npsh)]))
>         y <- (xdatu - u)/sc
>         y <- 1 + xi * y
>         if (min(sc) <= 0)
>             l <- 10^6
>         else {
>             if (min(y) <= 0)
>                 l <- 10^6
>             else {
>                 l <- sum(log(sc)) + sum(log(y) * (1/xi + 1))
>             }
>         }
>         l
>     }
>     x <- optim(init, gpd.lik, hessian = TRUE, method = method,
>         control = list(maxit = maxit, ...))
>     sc <- siglink(sigmat %*% (x$par[seq(1, length = npsc)]))
>     xi <- shlink(shmat %*% (x$par[seq(npsc + 1, length = npsh)]))
>     z$conv <- x$convergence
>     z$nllh <- x$value
>     z$vals <- cbind(sc, xi, u)
>     if (z$trans) {
>         z$data <- -log(as.vector((1 + (xi * (xdatu - u))/sc)^(-1/xi)))
>     }
>     z$mle <- x$par
>     z$rate <- length(xdatu)/n
>     z$cov <- solve(x$hessian)
>     z$se <- sqrt(diag(z$cov))
>     z$n <- n
>     z$npy <- npy
>     z$xdata <- xdat
>     if (show) {
>         if (z$trans)
>             print(z[c(2, 3)])
>         if (length(z[[4]]) == 1)
>             print(z[4])
>         print(z[c(5, 7)])
>         if (!z$conv)
>             print(z[c(8, 10, 11, 13)])
>     }
>     class(z) <- "gpd.fit"
>     invisible(z)
> }
>
>
> Kind regards,
>
> Char
>
>
> ________________________________
> Hotmail : un service de messagerie gratuit, fiable et complet Profitez-en
> ________________________________
> This automatic notification message was sent by Alliance
> (http://alliance.anu.edu.au/portal) from the Canberra R Users site.
> You can modify how you receive notifications at My Workspace > Preferences.
>



--
Felix Andrews / 安福立
Integrated Catchment Assessment and Management (iCAM) Centre
Fenner School of Environment and Society [Bldg 48a]
The Australian National University
Canberra ACT 0200 Australia
M: +61 410 400 963
T: + 61 2 6125 4670
E: felix.andrews-FCV4sgi5zeUQrrorzV6ljw@public.gmane.org
CRICOS Provider No. 00120C
-- --
http://www.neurofractal.org/felix/

This automatic notification message was sent by Alliance (http://alliance.anu.edu.au/portal) from the Canberra R Users site.
You can modify how you receive notifications at My Workspace > Preferences.