Comments on: How to draw simple graphs in R http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/ Thu, 11 Mar 2010 19:44:31 +0000 http://wordpress.org/?v=2.6.2 By: RAJASEKAR http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/#comment-29209 RAJASEKAR Sun, 28 Sep 2008 01:12:11 +0000 http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/#comment-29209 i'm a new user to matlab... i wanted basic progams..... i’m a new user to matlab… i wanted basic progams…..

]]> By: hongiiv http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/#comment-8555 hongiiv Wed, 09 Jan 2008 08:40:28 +0000 http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/#comment-8555 Hi, Bioinformatics Zen Blog is very useful for me. so I translate this blog's article to Korean(sure i'm Korean ^^;;). And don't want to, I will delete translated article. thank you. translate article here http://socmaster.homelinux.org/~hongiiv/299 Hi,

Bioinformatics Zen Blog is very useful for me. so I translate this blog’s article to Korean(sure i’m Korean ^^;;). And don’t want to, I will delete translated article. thank you.

translate article here http://socmaster.homelinux.org/~hongiiv/299

]]> By: Mike http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/#comment-2588 Mike Wed, 05 Sep 2007 12:57:46 +0000 http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/#comment-2588 @Kieren I use the MacOSX GUI for installingR packages. I'm not sure how to do them in linux. Try running install.packages("reshape") @Kieren
I use the MacOSX GUI for installingR packages. I’m not sure how to do them in linux.

Try running
install.packages(”reshape”)

]]> By: Kieren Lythgow http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/#comment-2570 Kieren Lythgow Tue, 04 Sep 2007 15:24:06 +0000 http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/#comment-2570 Hi Mike, Just working through some of your examples but when I try to install reshape it says the package isn't available. What command did you use to install it? Hi Mike,

Just working through some of your examples but when I try to install reshape it says the package isn’t available.

What command did you use to install it?

]]> By: Kanji distribution in common words at Kimtaro http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/#comment-1604 Kanji distribution in common words at Kimtaro Thu, 19 Jul 2007 10:43:21 +0000 http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/#comment-1604 [...] recently came across a nice introduction to The R Project for Statistical Computing, a nice way to create good looking statistical graphs. [...] [...] recently came across a nice introduction to The R Project for Statistical Computing, a nice way to create good looking statistical graphs. [...]

]]> By: ggplot: a plotting alternative to R base, and lattice http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/#comment-465 ggplot: a plotting alternative to R base, and lattice Mon, 14 May 2007 17:01:26 +0000 http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/#comment-465 [...] you found the tutorial on drawing graphs using R a bit of a kerfuffle, there’s a good introduction on drawing graphs using the ggplot package. [...] [...] you found the tutorial on drawing graphs using R a bit of a kerfuffle, there’s a good introduction on drawing graphs using the ggplot package. [...]

]]> By: Mike http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/#comment-462 Mike Mon, 14 May 2007 16:29:04 +0000 http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/#comment-462 Michele, thanks for your comments it's always nice to recieve any feedback. Michele, thanks for your comments it’s always nice to recieve any feedback.

]]> By: Mike http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/#comment-461 Mike Mon, 14 May 2007 16:27:08 +0000 http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/#comment-461 @Duncan and Neil Duncan does have a point. Why would you use R, a free piece of software, when there are enterprise applications such as Matlab and Mathematica? Especially as academics when we can usually buy them using an institution license. Well I'd say that Matlab and Mathematica are aimed at at mathematicians and physicists, and don't have the range of statistical functions that R does. Michael Crawley sums this up very well in the preface of <a href="http://eu.wiley.com/WileyCDA/WileyTitle/productCd-0470510242.html" rel="nofollow">The R Book</a>. Which I quote here <i> Why should you switch to R when you have mastered a perfectly adequate statistical package already. At one level there is no point in switching. If you only carry out a very limited range of statistical tests, and you don't intend to do more (or different) in the future, then fine. The main reason to switch to R is to take advantage of its unrivaled coverage and availability of its cutting edge applications in fields such mixed effects modelling and generalised additive models. </i> @Duncan and Neil
Duncan does have a point. Why would you use R, a free piece of software, when there are enterprise applications such as Matlab and Mathematica? Especially as academics when we can usually buy them using an institution license.

Well I’d say that Matlab and Mathematica are aimed at at mathematicians and physicists, and don’t have the range of statistical functions that R does.

Michael Crawley sums this up very well in the preface of The R Book. Which I quote here

Why should you switch to R when you have mastered a perfectly adequate statistical package already. At one level there is no point in switching. If you only carry out a very limited range of statistical tests, and you don’t intend to do more (or different) in the future, then fine. The main reason to switch to R is to take advantage of its unrivaled coverage and availability of its cutting edge applications in fields such mixed effects modelling and generalised additive models.

]]> By: Michele Mattioni http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/#comment-418 Michele Mattioni Sat, 12 May 2007 14:59:34 +0000 http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/#comment-418 Congratulations: Really good and clear. You've got one reader plus Congratulations:

Really good and clear.

You’ve got one reader plus

]]> By: Neil http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/#comment-392 Neil Fri, 11 May 2007 08:19:09 +0000 http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/#comment-392 Now Duncan. You know very well it's the other way around. Now Duncan. You know very well it’s the other way around.

]]> By: Duncan Hull http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/#comment-377 Duncan Hull Thu, 10 May 2007 11:09:32 +0000 http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/#comment-377 Hi Mike, nice post, and all useful stuff. Is R a bit like a poor mans Matlab? :) Hi Mike, nice post, and all useful stuff. Is R a bit like a poor mans Matlab? :)

]]> By: Mike http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/#comment-340 Mike Tue, 08 May 2007 12:00:59 +0000 http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/#comment-340 Thanks for the question, Doug. Regression is an in depth topic. It's simple to perform, but you need to keep in mind which assumptions you are making. The R method for non linear least squares in nls. You'll need to enter a formula, as well as a list of parameter estimates. There is also the selfStart method which automatically estimates the initial paramters. As for Monte Carlo analysis, this requires a little more coding. But could be applied for any type of regression. When you fit your model, you'll get a set of residuals, one for each point. This residual is effectively the distance between a point and the model prediction for that point. Randomise the list of residuals and use them to simulate a new set of data. I.e. Add the randomised set of residuals to the model predicted data. You can then refit your model to the new data set which will give you a new set of estimated parameters. Doing this process n-number of times will give you a set of estimated parameters, from which you could calculate the standard error for each. I'm sure there are better, and more statistically sound, ways of doing this, so I'd also recommend contacting the R <a href="https://stat.ethz.ch/mailman/listinfo/r-help" rel="nofollow">mailing list</a>. Thanks for the question, Doug.
Regression is an in depth topic. It’s simple to perform, but you need to keep in mind which assumptions you are making.

The R method for non linear least squares in nls. You’ll need to enter a formula, as well as a list of parameter estimates. There is also the selfStart method which automatically estimates the initial paramters.

As for Monte Carlo analysis, this requires a little more coding. But could be applied for any type of regression. When you fit your model, you’ll get a set of residuals, one for each point. This residual is effectively the distance between a point and the model prediction for that point. Randomise the list of residuals and use them to simulate a new set of data. I.e. Add the randomised set of residuals to the model predicted data. You can then refit your model to the new data set which will give you a new set of estimated parameters. Doing this process n-number of times will give you a set of estimated parameters, from which you could calculate the standard error for each.

I’m sure there are better, and more statistically sound, ways of doing this, so I’d also recommend contacting the R mailing list.

]]> By: Derek http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/#comment-329 Derek Mon, 07 May 2007 23:05:23 +0000 http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/#comment-329 thanks, for the nicely presented information thanks, for the nicely presented information

]]> By: doug http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/#comment-302 doug Sat, 05 May 2007 19:54:51 +0000 http://www.bioinformaticszen.com/2007/05/bioinformatics-simple-graphs-in-r/#comment-302 How about a tutorial on performing a non-linear least squares two parameter fit to an exponential equation, and a monte carlo simulation to estimate the error [this is something i have wanted to know how to perform in R recently] How about a tutorial on performing a non-linear least squares two parameter fit to an exponential equation, and a monte carlo simulation to estimate the error [this is something i have wanted to know how to perform in R recently]

]]>