Recently, I have been shopping for computers.
Anyone who are currently having direct control over more than 2 computers know the truth that it gets harder and harder to shop for another machine. There are just so many things to consider.
Maybe I should talk more about my current situation. I have a fairly powerful office workstation for working in my office, and I also have an okay powerful home desktop for evenings and weekends. Then I have a super light SONY VAIO laptop (3 pounds) for travel and teaching. Until recently, I was fairly happy with my arrangements. I have recently started my own religious schedule of backing up my office machines. I never leave important single copies on my home machine and laptop. All three machines share a virtual hard drive on our department's server (which won't be possible for anyone but for me it is possible since I use Columbia as my ISP at home). And Regina introduced me to this brilliant software--total commander, which compares directories and synchronizes them. I should be happy and productive.
But this summer, I need to go back China for nearly a month. And, God forbid, I am thinking about doing some work during my stay there. My little SONY VAIO baby is just not powerful enough for me to do serious work. It is still okay for powerpoint or LaTex (I maximized its RAM to 384MB) but it gets hot really fast and make extended use of it not quite enjoyable.
So I start thinking about getting another laptop. This just freaks me out. I don't quite like any of the windows laptop PC on market right now (picky, picky, picky). I think the Apple Powerbook looks cool and has a lot of neat applications. But thinking about managing 4 computers and 2 OS (occasionally I also need to do some UNIX stuff) is overwhelming. Am I getting too old? I dread the thought of remembering all the different shortcuts on different machines. That is fair. Maybe I will just get a window laptop then. But I can not even imagine installing all the essential softwares (even freely available) on this new computer. I wonder whether there is a way or a service that can duplicate the working style of my current machines and put it into a new (more powerful) machine. I suppose it is not the age then. It is my progressive laziness.
Tuesday, May 31, 2005
Thursday, May 26, 2005
So hard to grow up
Nearly everyday, I see some news about some tragic juvenile death (accidents or homocides). It really makes me sad and wonder what is the cumulated risks of dying before grown up. Here are some statistics on causes of death in 1998 I found online.
Sunday, May 22, 2005
Mean Absolute Deviation
Today, I tutored a talented young man on some high school statistics. It was about statistics on central tendency and dispersion, in a high school general mathematics book. Several things caught my attention.
1) The use of different terms. All the intro stat books we have examined for our teaching universally use "measure of center" and "measure of spread", not "central tendency" and "dispersion". Andrew once commented that only people for whom English is not a native language will find such usage of the language worthy of noting. Here, I am not fascinated by the differences. In my teaching, I have had a number of students who are confused by the different terms for a same thing in statistics. This is the first time I realize when such "variety" gets started.
2) MAD (mean absolute deviation) was taught. I just quickly checked several intro stat texts we use in this department. None of them cover MAD as a measure of spread. Only one covers a related topic, MAD regression. I can't understand what make such a topic essential to be in a high-school math text but unnecessary in a college intro stat text. Personally, I like MAD.
3) The extensive use of TI-83/84 calculator. It is very foreign to me for a math textbook to illustrate how to use a calculator. Using calculators for homework or exams weren't acceptable when I was in China. The use of calculator has the advantage of enabling the students to work on real-life scale of statistical problems. But following the diagram such as "stats->2->enter->..." In the book, the student, as my pupil said today, "had no idea what he was doing."
This is not an AP stat course. To my understanding, the text used for AP stat is usually very similar to college text but the emphasis is more on the computation not too much on the concepts.
1) The use of different terms. All the intro stat books we have examined for our teaching universally use "measure of center" and "measure of spread", not "central tendency" and "dispersion". Andrew once commented that only people for whom English is not a native language will find such usage of the language worthy of noting. Here, I am not fascinated by the differences. In my teaching, I have had a number of students who are confused by the different terms for a same thing in statistics. This is the first time I realize when such "variety" gets started.
2) MAD (mean absolute deviation) was taught. I just quickly checked several intro stat texts we use in this department. None of them cover MAD as a measure of spread. Only one covers a related topic, MAD regression. I can't understand what make such a topic essential to be in a high-school math text but unnecessary in a college intro stat text. Personally, I like MAD.
3) The extensive use of TI-83/84 calculator. It is very foreign to me for a math textbook to illustrate how to use a calculator. Using calculators for homework or exams weren't acceptable when I was in China. The use of calculator has the advantage of enabling the students to work on real-life scale of statistical problems. But following the diagram such as "stats->2->enter->..." In the book, the student, as my pupil said today, "had no idea what he was doing."
This is not an AP stat course. To my understanding, the text used for AP stat is usually very similar to college text but the emphasis is more on the computation not too much on the concepts.
Friday, May 20, 2005
Modeling DNA Base Substitution in Large Genomic Regions from Two Organisms
J Mol Evol. 2004 Jan;58(1):12-8.
Von Bing Yap and Terence P. Speed
Abstract: We studied the substitution patterns in 7661 well-conserved human–mouse alignments corresponding to the intergenic regions of human chromosome 22. Alignments with a high average GC content tend to have a higher human GC content than mouse GC content, indicating a lack of stationarity. Segmenting the alignments into four groups of GC content and fitting the general reversible substitution model (REV) separately gave significantly better fits than the overall fit and the levels of fit are close to that expected under an REV model. In addition, most of the fitted rate matrices are not of the HKY type but are remarkably strand-symmetric, and we constructed a number of substitution matrices that should be useful for genomic DNA sequence alignment. We did not find obvious signs of temporal inhomogeneity in the substitution rates and concluded that the conserved intergenic regions in human chromosome 22 and mouse appear to have evolved from their common ancestors via a process that is approximately reversible and strand-symmetric, assuming site homogeneity and independence.
Von Bing Yap and Terence P. Speed
Abstract: We studied the substitution patterns in 7661 well-conserved human–mouse alignments corresponding to the intergenic regions of human chromosome 22. Alignments with a high average GC content tend to have a higher human GC content than mouse GC content, indicating a lack of stationarity. Segmenting the alignments into four groups of GC content and fitting the general reversible substitution model (REV) separately gave significantly better fits than the overall fit and the levels of fit are close to that expected under an REV model. In addition, most of the fitted rate matrices are not of the HKY type but are remarkably strand-symmetric, and we constructed a number of substitution matrices that should be useful for genomic DNA sequence alignment. We did not find obvious signs of temporal inhomogeneity in the substitution rates and concluded that the conserved intergenic regions in human chromosome 22 and mouse appear to have evolved from their common ancestors via a process that is approximately reversible and strand-symmetric, assuming site homogeneity and independence.
Bayesian Phylogenetic Model Selection Using Reversible Jump Markov Chain Monte Carlo
Mol. Biol. Evol. 21(6):1123-1133. 2004
John P. Huelsenbeck, Bret Larget, and Michael E. Alfaro
Abstract: A common problem in molecular phylogenetics is choosing a model of DNA substitution that does a good job of explaining the DNA sequence alignment without introducing superfluous parameters. A number of methods have been used to choose among a small set of candidate substitution models, such as the likelihood ratio test, the Akaike Information Criterion (AIC), the Bayesian Information Criterion (BIC), and Bayes factors. Current implementations of any of these criteria suffer from the limitation that only a small set of models are examined, or that the test does not allow easy comparison of non-nested models. In this article, we expand the pool of candidate substitution models to include all possible time-reversible models. This set includes seven models that have already been described. We show how Bayes factors can be calculated for these models using reversible jump Markov chain Monte Carlo, and apply the method to 16 DNA sequence alignments. [Follow the link above to read more.]
John P. Huelsenbeck, Bret Larget, and Michael E. Alfaro
Abstract: A common problem in molecular phylogenetics is choosing a model of DNA substitution that does a good job of explaining the DNA sequence alignment without introducing superfluous parameters. A number of methods have been used to choose among a small set of candidate substitution models, such as the likelihood ratio test, the Akaike Information Criterion (AIC), the Bayesian Information Criterion (BIC), and Bayes factors. Current implementations of any of these criteria suffer from the limitation that only a small set of models are examined, or that the test does not allow easy comparison of non-nested models. In this article, we expand the pool of candidate substitution models to include all possible time-reversible models. This set includes seven models that have already been described. We show how Bayes factors can be calculated for these models using reversible jump Markov chain Monte Carlo, and apply the method to 16 DNA sequence alignments. [Follow the link above to read more.]
Subscribe to:
Posts (Atom)