Bayesian data analysis recommendations

[tmrsg11]

Dear all,
I am trying to learn Bayesian inference and Bayesian data analysis, I
am new in the field.  Would any experts on the list recommend any good
sites or materials for beginners?

My approach is to learn and understand the theory first, then program
on my own using R, though I see there are already packages.

appreciate any help, thanks in advance!

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.

[Rich Shepard]

On Thu, 19 Jan 2012, C W wrote:

> I am trying to learn Bayesian inference and Bayesian data analysis, I am
> new in the field.  Would any experts on the list recommend any good sites
> or materials for beginners?
>
> My approach is to learn and understand the theory first, then program
> on my own using R, though I see there are already packages.

   I'm far from an expert, but why not avoid re-inventing the wheel while you
learn? Buy and read Jim Albert's "Bayesian Computation with R".

   If you're a population ecologist (or willing to extend pesented examples
and ideas to communities and ecosystems), Ben Bolker's "Ecological Models
and Data in R" explains when Bayesian and frequentist approaches each have
advantages over the other.

Rich

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.

[tmrsg11]

Thanks, Rich, I will look at the book.

I agree, there are many nice packages, but what if the package changes in a
few years?  I would have no idea what is going on!  I've heard
from predecessor in the industry who emphasize the learning, not just plug
and chug.

I really want to learn the material and understand it, above all, it is
interesting.

I am looking more towards Bayesian statistics or Bayesian inference.  I am
in statistics graduate school, though not my field, the biology application
could help in the understand I suppose?

On Thu, Jan 19, 2012 at 7:07 PM, Rich Shepard <[hidden email]>
wrote:

> On Thu, 19 Jan 2012, C W wrote:
>
>> I am trying to learn Bayesian inference and Bayesian data analysis, I am
>> new in the field.  Would any experts on the list recommend any good sites
>> or materials for beginners?
>>
>> My approach is to learn and understand the theory first, then program
>> on my own using R, though I see there are already packages.
>
>
>  I'm far from an expert, but why not avoid re-inventing the wheel while

you

> learn? Buy and read Jim Albert's "Bayesian Computation with R".
>
>  If you're a population ecologist (or willing to extend pesented examples
> and ideas to communities and ecosystems), Ben Bolker's "Ecological Models
> and Data in R" explains when Bayesian and frequentist approaches each have
> advantages over the other.
>
> Rich
>
> ______________________________________________
> [hidden email] mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide

http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.

        [[alternative HTML version deleted]]

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.

[Rich Shepard]

On Thu, 19 Jan 2012, C W wrote:

> I agree, there are many nice packages, but what if the package changes in
> a few years?  I would have no idea what is going on!  I've heard from
> predecessor in the industry who emphasize the learning, not just plug and
> chug.

> I really want to learn the material and understand it, above all, it is
> interesting.

   You'll learn the underlying theory from these books; the R code is the
chosen medium for examples that illustrate the theory.

   In the F/OSS world when packages are released in new versions they are
either backwards compatible (to prevent breaking existing applications) or
they provide plenty of notice of incompatibility. You can also read the
source to see how it works and run diff on the sources to pick up
diffenrences.

> I am looking more towards Bayesian statistics or Bayesian inference.  I am
> in statistics graduate school, though not my field, the biology
> application could help in the understand I suppose?

   If you're a mathematical statistician the biological/ecological examples
may or may not be of value to you. It all depends on what you plan to do
with your degree.

Rich

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.

[Marc Schwartz-3]

On Jan 19, 2012, at 6:07 PM, Rich Shepard wrote:

> On Thu, 19 Jan 2012, C W wrote:
>
>> I am trying to learn Bayesian inference and Bayesian data analysis, I am
>> new in the field.  Would any experts on the list recommend any good sites
>> or materials for beginners?
>>
>> My approach is to learn and understand the theory first, then program
>> on my own using R, though I see there are already packages.
>
>  I'm far from an expert, but why not avoid re-inventing the wheel while you
> learn? Buy and read Jim Albert's "Bayesian Computation with R".
>
>  If you're a population ecologist (or willing to extend pesented examples
> and ideas to communities and ecosystems), Ben Bolker's "Ecological Models
> and Data in R" explains when Bayesian and frequentist approaches each have
> advantages over the other.
>
> Rich

Another reference would be:

Doing Bayesian Data Analysis by Kruschke
http://doingbayesiandataanalysis.blogspot.com/

It is based upon the use of R and BUGS, but was just updated so that the download of the book's code examples also includes the use of JAGS code, which is more easily used on OS's other than Windows.

HTH,

Marc Schwartz

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.

[Matt Shotwell-2]

On Thu, 2012-01-19 at 19:23 -0500, C W wrote:

> Thanks, Rich, I will look at the book.
>
> I agree, there are many nice packages, but what if the package changes in a
> few years?  I would have no idea what is going on!  I've heard
> from predecessor in the industry who emphasize the learning, not just plug
> and chug.
>
> I really want to learn the material and understand it, above all, it is
> interesting.
>
> I am looking more towards Bayesian statistics or Bayesian inference.  I am
> in statistics graduate school, though not my field, the biology application
> could help in the understand I suppose?

This list (r-help) may not be the best place to look for advice on this.
But here is some anyway :)

For a well-rounded introduction, I recommend Robert's 'The Bayesian
Choice'. This is a great foundation for Bayesians who intend to defend
their positions on statistical inference. For a more practical approach,
Gelman, Carlin, Stern, and Rubin's book 'Bayesian Data Analysis' has
been very popular (THE most popular, according to some). Regarding the
software tools for Bayesian data analysis, the most mature _and_ active
_and_ best integrated with the R project is Martyn Plummer's JAGS (See
also the R package rjags, by the same author). Another tool that I'm
planning to check out is PyMC: http://code.google.com/p/pymc/

Best,
Matt

> On Thu, Jan 19, 2012 at 7:07 PM, Rich Shepard <[hidden email]>
> wrote:
> > On Thu, 19 Jan 2012, C W wrote:
> >
> >> I am trying to learn Bayesian inference and Bayesian data analysis, I am
> >> new in the field.  Would any experts on the list recommend any good sites
> >> or materials for beginners?
> >>
> >> My approach is to learn and understand the theory first, then program
> >> on my own using R, though I see there are already packages.
> >
> >
> >  I'm far from an expert, but why not avoid re-inventing the wheel while
> you
> > learn? Buy and read Jim Albert's "Bayesian Computation with R".
> >
> >  If you're a population ecologist (or willing to extend pesented examples
> > and ideas to communities and ecosystems), Ben Bolker's "Ecological Models
> > and Data in R" explains when Bayesian and frequentist approaches each have
> > advantages over the other.
> >
> > Rich
> >
> > ______________________________________________
> > [hidden email] mailing list
> > https://stat.ethz.ch/mailman/listinfo/r-help
> > PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> > and provide commented, minimal, self-contained, reproducible code.
>
> [[alternative HTML version deleted]]
>
> ______________________________________________
> [hidden email] mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.

[H. T. Reynolds]

You might look at John Kruschke's book, Doing Bayesian Data Analysis (AP), which starts with basics and goes from there. It also relies on R and Bugs.

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.

[wibeasley]

Even if you're not doing medical research,  I like a lot about Spiegelhalter's book:
http://www.amazon.com/Bayesian-Approaches-Health-Care-Evaluation-Statistics/dp/0471499757/ref=sr_1_1?ie=UTF8&qid=1327112075&sr=8-1

For interacting with R and JAGS/BUGS my two favorite books that cover theory are Carlin & Louis and the 2nd half of Gelman & Hill.
http://www.amazon.com/Bayesian-Methods-Analysis-Chapman-Statistical/dp/1584886978
http://www.amazon.com/Analysis-Regression-Multilevel-Hierarchical-Models/dp/052168689X

If you have a handle on the theory, Jim Ablert's book (previously mentioned by Rich Shepard) is fun.

[yvonnick noel]

Hi,

On the R side, you may want to have a look at the AtelieR package. It's
a GTK GUI which gives you a simple interface to some common Bayesian
tests (on a proportion, on a variance, on a mean, on mean and variance
jointly, on several proportions, on contingency tables, on several means).

There are also some automatic search procedures of the best model, when
comparing several means, proportions, or rows in a contingency table.

Hope this may be useful,

Yvonnick Noel
University of Brittany
Department of Psychology
Rennes, France

> Dear all,
> I am trying to learn Bayesian inference and Bayesian data analysis, I
> am new in the field.  Would any experts on the list recommend any good
> sites or materials for beginners?
>
> My approach is to learn and understand the theory first, then program
> on my own using R, though I see there are already packages.
>
> appreciate any help, thanks in advance!
>

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.