Using R for stats (would prefer Excel)

[jrohmann]

A class I am taking encourages using R for stats but, since I know and use Excel all day, daily, I would prefer to use it - but am not sure if I can do most of the same things. Anyone know? Those who use R love it and advise against using Excel for this. Thanks!

 

[shg]

Excel takes a lot of heat for it's statistical functions, but they've gotten significantly better in every version since 2003.

 

[AhoyNC]

I have used R for about a year now. Both are about equal for basic statistics, but I find R to be better for advanced statistics and a big improvement for graphing over Excel. R does have a steep learning curve and as you have found out does not have a nice GUI interface. However once I learned the basics of R I now prefer it over Excel for statistical work. If you are just taking a basis stat class you should be able to use Excel. I've found R to be better and easier than Excel at making histograms and running correlations.
If your course requires R then probably you will struggle using Excel for assignments.

These web sites may be of help with basis R.
twotorials by anthony damico
http://faculty.washington.edu/tlumley/Rcourse/R-fundamentals.pdf
Good luck.

 

[jrohmann]

Thank you!

 

[AhoyNC]

You're welcome.
You may also find these web sites helpful
R Tutorials--Data Frames


For graphs:
http://www.harding.edu/fmccown/r

 

[jrohmann]

I saw all of those functions and those I've used seemed easy and matched with R so I wondered why so many think excel is so much worse, so far anyway. Have done histograms in both and for that I prefer Excel so will probably continue to learn in both and compare. I am very biased toward excel tho.

 

[AhoyNC]

I don't think Excel is worse. In fact like you I didn't care for R when I was forced to use it. I still use Excel to clean data before downloading it to R.
One example I can think of is try to do a box plot in Excel. It takes a lot of work in Excel and one simple command in R. For basic stat. Excel and R are about equal in my opinion.
Remember R is a stat. program and you can find 100's of stat. packages in R, which you would have to write yourself if using Excel. Once you get use to writing R scripts things get easier. If you are only going to be doing things like averages, standard dev., scatter plots, etc., I would stick with Excel. Otherwise I think you will find R is more versatile than Excel.
R can be a very frustrating program to learn, but it goes get easier as you start to learn the commands.

 

[jrohmann]

I finally got a histogram to work by cleaning the data in excel first, like you do, so maybe I am on the right track. However, I don't even know what a package is. I only know how to load them and that you have to do library() on it before you can use it. This is very humbling. I will keep looking for "You just downloaded R, now what" tutorials!
This class started it's instruction from somewhere beyond pure beginner and many are having trouble because we don't know enough of the basics. Even reading ?hist was overwhelming. But it is what it is and if I get stuck at least I have excel to go to for now.

 

[AhoyNC]

You are right in the help and error msgs. in R are useless to a beginner in most cases. Packages are just what you download for the libraries you refer to. You can find them for QC charts, analyzing stocks, etc. If you haven't looked at it the cran website has a good intro to R.
http://cran.r-project.org/doc/contrib/Verzani-SimpleR.pdf

If you have the time to stick with it, I think you will find R is a good program.

 

[jrohmann]

This looks great and very helpful. Thanks!