R on the iPad: what we've got so far

I have just received an iPad as a Christmas present from my beloved husband. And there is a question I have already raised concerns about, that is the compatibility of R with the iOS devices. It is proven that R language is successfully compiled by the iPhone development tools, unfortunately, according to the Apple's Terms of Service, no GPL software can be acquired through the App Store. Hereby, either we have to wait till Apple modifies its TOS, or - jailbreak our iPad/iPhone. Then there is a completely described procedure of installing R on the iOS (jailbroken) devices.

Meanwhile, a web interface for R can be used since the access to internet is available: the RWeb compiles the R code once entered in a text box and returns the output on an html-page. Although the process goes a little slow, we cannot but consider the fact the plotting options are also available (as well as the year of creation of the service: 1998!)

As for Apple App Store, the only "R application" available by now is the R-Chart, the news service of the R language programming community.

However, we hope Apple would reconsider its decision over the R client for its iOS devices - due to the high and spreading demand both on R and mobility.

I personally would be very glad to receive any comments, corrections, news, updates, and hints. Happy holidays everyone!

"Oh, no!", a hint and a quiery

Remember the superbook on Spatial Statistics? The guy who wrote it (co-), gave a talk in the Universitat Politécnica de Catalunya last Friday. And... your obedient servant was late there while drying here hair at home and looking for a parking spot. NB: Be punctual!

Then, I have a funky dataset of 400 obs and trying to make it more...er... vivid! The point is that I have a Mac, and my only spreadsheet application is the "Numbers" from the iWork package. As far as I know it, it saves data only in its internal format that has a ".numbers" what-a-large extension, or in ".xls". Forget the ".csv"s, ".dbf"s and all extensions you are used to. The data we don't compile, however (God bless the databases administrators!), we receive it as it is. And, if your ".csv"-data is er... laconic you could make it more vivid by making some data mining on your own and adding what you found to your set. For doing that with Mac you have to be able to open your ".csv" as a spreadsheet. You could do that using "Google Docs" by uploading the data, editing it online, and then downloading the vivid dataset as what you want. But you fully depend on your internet connection and the size of your set. Another option is that you use the R command write, save the data as an ".xls"-file, edit it in your "Numbers", then save the result as an ".xls" and read it to R as a new object. Or, if the data is big, you can download one variable, add the data to it in the corresponding order (so you don't have to further sort) and then, read the result to R and merge the two datasets by a key.

Do you have your less lamer tricks to add data to the existing dataset in R? Directly?

Thanks, and don't miss interesting talks!

Lost, lost, lost!

I haven't written for a while, because my self-organising module was on vacation, i.e. out of order.

Now in September we are all seriously back to work (in Spain it is even called "Operación Setiembre", no translation needed). I have entered in a new field of research, and it is oh-so-spatial, and as far as I barely have any clue of how the spatial data is treated, I stick to a perfect-perfect-perfect book:



Here you can read it online. I have a hard copy from the University's Lib, he-he.

But my problems start when it comes to practice. As I try to install required packages to train new methods, my fancy R editor for Mac OS just hangs!!! Might be, this is due to slow internet connection at the lab, but hardly.

I need a fancier R editor for Mac OS! Any suggestions?

Use R! 2010 conference starts today

I think you understand the message I try to bring home to you through this image. Though this year the majority of R users is not presented at the conference, it's a nice challenge for the upcoming paper to do some nice research with/on R, write some nice paper about it, and participate in the conference. See you there next year!

And by now all we can and should do is track this conference via internet:

user2010.org

A following strongly readable blog entry talks about the statistics of the conference through the last years, with a code provided for one to program/plot it all by oneself:

R-Chart: Analyze Online R User Conference Data

Friday weirdo

Though I promised myself I would post twice a week in order to increase self-discipline and track my work flow, when it comes from words to their execution it somehow becomes incredibly hard...

Anyway, this nice and sunny Münchner Friday I would like to remind some of those who try artistic plotting with R that the whole range of colors is available by typing colors().

I did try artistic plotting and loved the following ones from the default pattern: "turquoise3", "sienna1" and "yellowgreen" (not to be mixed up with "greenyellow"!!!!!!!!!!!! :-)):



The data is taken from the Institut d'Estadistica de Catalunya (IDESCAT).

However, if defaults do not satisty your thirst for art, you can always reach all the RGB palette.

Have nice plots!

tweet with R

Pals,

I cannot load the twitteR package for R. After reading this brilliant entry on "R bloggers", I crave to do something alike, but fail to set the machine up.

I have downloaded the package and tried to call it, but R doesn't give me the green light. It says:

> library(twitteR)
Loading required package: RCurl
Error: package 'RCurl' could not be loaded
Además: Mensajes de aviso perdidos
In library(pkg, character.only = TRUE, logical.return = TRUE, lib.loc = lib.loc) :
there is no package called 'RCurl'

and when I look up this RCurl (which I have no idea what does) and try to load it manually, R says:

ERROR: dependencies ‘bitops’ are not available for package ‘RCurl’
* removing ‘/Users/lyolya/RCurl’

What shall I do with that, poor stupid me???

Disclaimer

As I am not able to regularly update the content here, and this blog is valuable to me, I have decided to give it a little more value by transforming it into my personal scientific monitor.

I give myself a command to publish a new entry here at least twice a week in order to see what I have learned, what I have discovered, and how the things are going on in general.

Now I am in Germany, realizing my doctoral practice in Munich. At work I deal with the proportional odds statistical model and make a lot of use of the R software and of its VGAM package in particular. Everything is new to me, due to the fact I am working with Mac since recently. This is my first Mac, and to me it's freaky! I hope I'll get over it soon, and we'll make friends.

Everybody who'll ever use VGAM for proportional odds model prease bear in mind two things:

- ...family=propodds doesn't really work; use the family=cumulative(parallel=TRUE) instead;
- while interpreting the model's summary remember that VGAM accords to the following approach:

logit[P(Y<=y|x)]=a - bx

so the coefficient's impact should be consumed with the opposite sign.

VGAM is tricky but it is a really good stuff.