Skip to main content

Elementary analysis resources

8 posts / 0 new
Last post
Matthew Templeton
Elementary analysis resources

Hi,

One of the things that I think we could use here is a list of things that people might find useful when doing analysis of any sort.  I decided to start a thread on it and encourage people to add suggestions.  If we get a really comprehensive list, we can create a more centralized webpage with the information.

I think I'll start not with information about analysis techniques, but about how to find out things about a star or data set that you're interested in analyzing.  I use a lot of resources on a regular basis, but I'm not sure whether they're generally known about.

Probably your first place to stop is VSX itself, since it's very comprehensive and there are a lot of people working very hard to make it even more so.  Remember you can do both name and RA/Dec searches, so you can find fairly quickly what's known or not known about an object.  The URL is http://www.aavso.org/vsx

There are also a number of academic and professional sites that are really helpful, with the big five for me being:

  • ADS, for general literature searches
  • arXiv.org, for (some) information about what's been published recently
  • Simbad, for literature on specific stars or other objects
  • VizieR, for data searches on objects and fields, and
  • Skyview, for quick, multiwavelength views of fields and coordinates

There are a number of different phrasings, but there's a good humorous saying to the effect of "a month at the telescope can save you an hour in the library".  It's always fun to observe, and it's also fun to look at data, but one of the best ways to start a data analysis project is to learn what's already known about the objects you're looking at.  There are a lot of data analysis tasks that can be done very quickly (like running a Fourier transform in VStar), but be aware the body of existing literature and information for some stars is extremely large.  It's good to know what previous work has and hasn't said before you start a really comprehensive analysis program.

Feel free to start adding ideas and other useful sites that you've encountered.

Matt

[edited, changed "VPhot" to "VStar"]

Aaron Price
Aaron Price's picture
more intro resources

Also, the Data Analysis web page has a introductory resources for the analysis side of things:

David Benn
David Benn's picture
Adding to Matt and Aaron's

Adding to Matt and Aaron's lists:

  • The AAVSO in Print is a source of information about literature derived from AAVSO-data/resources. I learned about this from...
  • CHOICE: Variable Star Classification and Light Curves, a worthwhile course to begin thinking systematicaly about variable star types and light curves.
  • Grant Foster's book: Analyzing Light Curves: a Practical Guide.

Matt:

  • I like the "... a month at the telescope..." quote. It reminds me of this one (Thomas Sowell):

It takes considerable knowledge just to realise the extent of your own ignorance.

  • You may have meant "... like running a Fourier transform in VStar" above? :P

David

Matthew Templeton
VPhot != VStar

David,

Oops, yes, I've had VPhot on my mind recently with the CCD Photometry course.  Post is edited -- thanks for the catch!

Matt

marlin
marlin's picture
Comprehensive R software

I bought Grant Foster's book on light curves and it requires Comprehensive R software.  This is more dense that the mathematics needed to follow the arguments.  Is there a more user friendly software that will work with the book or a dummy's guide to the R?  Ireally lilke the subject but don't want to spend a semester just figuring out the program.

Marlin Costello

David Benn
David Benn's picture
Foster, R

Hi Marlin

I guess it depends upon what you want to achieve.

Using RStudio will make it slightly simpler to work with R code but it's still going to require you to understand the code. There are many resources on the web that can help, but some effort is required of course, e.g. 

If you want to use a different programming language, the problem you'll face is the need to understand that language and port the R code from Foster's book to it.

If you want to do light curve analysis based upon principles in Foster's book, you could use software (like VStar).

I'm not sure whether this helps.

David

Eric Dose
Eric Dose's picture
R, RStudio, getting started

R is used mostly as a trial-and-error language and framework, best for statistics and vector math (e.g., for time series, spectra, etc). Most R users type in a line at a time. If the typing gets too repetitive, it's time to package up the repeated lines into a function (easy) and then include the function's name (and maybe parameter values) in new typed-in lines.

R is exploratory. It's not really meant for architected programs, as you'd use Java, C, etc. It has a slightly weird syntax, but it was written by statisticians after all (and to be fair, it's not nearly as weird as its primeval ancestor, APL). The fastest (and most fun!) way to get comfortable with R is to type interesting stuff into RStudio's "console" window and see what happens. You can't break it.

R and RStudio and all packages are free. Vectors, data frames, and a lot of statistics are already in the base R installation. For more specific needs, package installs are much easier than for Python, etc.

RStudio is by far the most organized way to use R, for almost all purposes. I wouldn't use R without it. It doesn't change the R lines you type in at all, but it does make it much easier to write small scripts, repeat calculations, store and retrieve files, etc.

Geoff
R programming

I took this free course offered through Coursera a while back.  It's a basic introduction to R.  Coursera also has a number of other courses using R for statistics and more generally data science.  

https://www.coursera.org/learn/r-programming

 

Log in to post comments
AAVSO 49 Bay State Rd. Cambridge, MA 02138 aavso@aavso.org 617-354-0484