Friday, April 30, 2010

Ubuntu 10.04 cursor issue

I am having the same problems as these guys. The workaround seems to be to disable the desktop effect first, change the cursor, then enable desktop effect back. Hope to have a more convenient way way soon.

One good thing with the new release is that the annoying sound comes when the login screen shows up can be disabled. I tried many ways but still could not disable it with 9.10.

Upgrade to Ubuntu 10.04

Just upgraded my Ubuntu box from 9.10 to 10.04. The process went all right, the overall system performance seems satisfactory, but I did not like the change to move the window controls from upper right side to the upper left side.

And I am still using the Clearlook theme.

JAGS for Bayesian analysis

One of the many reasons I like Jackman's book "Bayesian Analysis for the Social Sciences" is that it relies on JAGS as the primary computational platform (as opposed to WinBUGS). From what I know, it is the first Bayesian book that does this.

Thursday, April 29, 2010

Wednesday, April 28, 2010

Sweave vs. pgfSweave

I think pgfSweave should replace the default of Sweave system because it brings significant improvements over the original Sweave. I certain have made it my default Sweave tool on my system.

------------------------------- EDITED ON MAY 1, 2010 --------------------------------------

Since StatET made Sweaving so simple and easy, I don't have incentive to work directly on LaTeX documents any more. It is actually much simpler and more intuitive to work directly on a Sweave documents and see results in the final PDF output.

Tuesday, April 27, 2010

StatET does not work with OpenJDK

After several failed attempts to get StatET to work on my older Ubuntu box, I realized that it does not work with OpenJDK. After replacing the default OpenJDK with the official Sun JDK, everything works fine.

Spell checking in Eclipse

This post explains how to improve spell checking in Eclipse.

The new JAGS is out

Version 2.0 of the cross-platform BUGS language implementation is out.

Word wrap in eclipse

For those who struggle to get reasonable word wrapping function in Eclipse, this is a good news.

Here is the original source.

Friday, April 23, 2010

Trouble with ESS and Sweave

Last time I tried to sweave a document from with Emacs+ESS, I was using an earlier version of ESS (the current version is 5.8), and things seemed to be fine. Today when I tried to sweave a simple document and produced PDF output, I got error message of ".pdf": exited abnormally with code 2". I posted a message to the ESS mailing list, but so far no solutions have been reported. 


Here is a very simple .Rnw file I used to reproduce the error: 

---------------------------------------
\documentclass[12pt]{article}

%\usepackage{tikz}
%\usepackage{fullpage}
\usepackage{Sweave}

\begin{document}

This document outlines the analysis to be carried out to look at the
relationship between prenatal famine exposure and fetal loss risk.

\end{document}
-----------------------------------------
Before this can be fixed, I will rely on Eclipse for all my sweaving. 
---------------- April 26 ------------------------
Here is my own solution: 

from line 177 of ess-svy.el:

-      (shell-command (concat -       (if (and ess-microsoft-p (w32-shell-dos-semantics)) -           "start \"" pdfviewer "\" \"" namestem ".pdf\"" -                 "\"" pdfviewer "\" \"" namestem ".pdf\" &")))) +      (shell-command (concat
+                 "\"" pdfviewer "\" \"" namestem ".pdf\"")))

Rumour: Apple To Acquire ARM?

I believe that this, if not a rumor, is against most people's interest, manufactures and consumers alike.

Tuesday, April 20, 2010

Processing

This is a nice little toy.

Monday, April 12, 2010

Rampant cheating hurts China's research ambitions

This is interesting. Smells almost like the Great Leap Forward.

Friday, April 09, 2010

GLMM using DPpackage

I was able to fit a semi-parametric Bayesian GLMM model using DPpackage. It took me many hours to sample from the posterior distribution (DPM prior):

MCMC scan 1000 of 5000 (CPU time: 18950.080 s)
MCMC scan 2000 of 5000 (CPU time: 22510.100 s)
MCMC scan 3000 of 5000 (CPU time: 28293.830 s)
MCMC scan 4000 of 5000 (CPU time: 35111.930 s)
MCMC scan 5000 of 5000 (CPU time: 46726.330 s)

Which translates to 5.26, 6.25, 9.75, 12.98 hours. This makes it less suitable for routine (especially exploratory) data analysis.

I compared the results from DPpackage and that from MCMCglmm, and they are not that different, and the latter took only a small fraction of the time required by the former!

The lack of difference in results puzzled me. I compared from results from random effect logistic regression assuming Gaussian random effect and results from NPML, assuming a nonparametric distribution of the random effect, the differences are quite significant.

---------------------------- UPDATED ON APRIL 11 ----------------------------------------------------------------

Using DP prior instead of DPM prior, it took about 4.7 hours to run the model. The results are slightly different and the parameter I am interested in increased from .41 to .42. Now I am trying PT prior and see how it goes.

DPpackage is a exciting new tool for applied researchers, and A LOT OF new and cool things can be done with it. With convenient new Bayesian tools like MCMCpack, MCMCglmm, and DPpackage, I will not be surprised to see more Bayesian publications coming out in social sciences. 

Tuesday, April 06, 2010

Emacs color theme

Here is an emacs color theme gallery.

Five open source alternatives to the iPad

Here you can find some alternatives to Apple's ipad, running on Linux.

12 More of the Best Free Linux Books

Very good and free books on Linux can be found here.

OpenGeoDa

The new version of the fabulous open source GIS software, GeoDa, called OpenGeoDa, has a Linux version as well as win32 and apple version. I have not tried it, but this is really a great news!

Monday, April 05, 2010

Sunday, April 04, 2010

A Practical Guide to Geostatistical Mapping

Good book and good introduction to geostatistical modeling.

Friday, April 02, 2010

BAT: Bayesian analysis toolkit

BAT is a Bayesian analysis software. Looks like it needs quite some programming to get anything useful done.

----------------------- added on April 4 --------------------------

I seem to have troubles compiling it on my 64-bit Ubuntu box.

Thursday, April 01, 2010

Apple got slapped

According to this post, Apple does not seem to be the first one to have the idea of "multitouch".

Counter