The Museum of HP Calculators

HP Forum Archive 15

[ Return to Index | Top of Index ]

Stat49Pro+ Impressions Please...
Message #1 Posted by Scott Guth on 7 Oct 2005, 8:31 p.m.

Hello all,

Stat49Pro+ and Stat48Pro+ ver. 3.0 will be available for the HP49G(+) and HP48G(X) calculators soon. In case you are unfamiliar with these products, a description of their capabilities is given below.

What I need is a few people who are familiar with basic and hopefully more advanced methods of inferential statistics who would be willing to try out the product and send suggestions. The HP48G(X) and HP49G(+) versions are both available for this.

If you are interested in previewing the product or seeing a detailed data sheet on the product, please send me an email at guthjun...@charter.net.

Thanks,

Scott Guth Mathematics Dept., Mt. San Antonio College

What's new in Stat49Pro+?

1. Arbitrary regression/least squares applications with inference methods. a. Linear regression with inferences on slope, rho, expected/predicted y values. b. Multivar-linear regression c. Polynomial regression d. Multivar non-linear regression e. Many tests & confidence intervals for these applications

2. Better integration with the Data Manager. The Data Manager can now be invoked from within any inference application - just press the [Data] soft-menu key, and enter your data. Once the data is stored, press the [ImPrt] soft-menu key to import whatever statistics are needed from your data set directly into your input form.

3. More ways to Describe multivariate statistical data (in the Data Manager), including Mean Vector, Variance/Covariance Matrix, and the Correlation Matrix.

4. Better inferential methods for normal-quantile plots.

The following gives an overview of the older features of the product:

Stat49Pro+ is a complete inferential statistics library written in System RPL. Essentially all methods of descriptive and inferencial statistics covered in a college level statistics curriculum are computed by Stat49Pro+. Included is a data manager which stores sample data and sends statistics directly into input forms for the various inference applications. Some non-parametric tests are supported as well, along with normal probability plots and rank sums.

REPLACE THIS TEXT WITH YOUR LISTING

      
Re: Stat49Pro+ Impressions Please...
Message #2 Posted by Scott on 7 Oct 2005, 8:41 p.m.,
in response to message #1 by Scott Guth

Sorry, my email is guthjunk77@charter.net -Scott Guth

            
Better Short Median Formula
Message #3 Posted by I, Claudius on 7 Oct 2005, 10:53 p.m.,
in response to message #2 by Scott

Aha, someone who knows his stats! One of the missing functions in the 48/49 stat menu is a median function. I wrote a User RPL program that uses the short formula of median position = 1/2*(n+1),

<< SORT DUP SIZE 1 + 2 / GET LASTARG FLOOR GET + 2 / >>

but it obviously won't work in all cases. Is there a better short formula for the median I can use?

Edited: 7 Oct 2005, 11:06 p.m.

                  
Re: Better Short Median Formula
Message #4 Posted by Scott on 10 Oct 2005, 12:54 a.m.,
in response to message #3 by I, Claudius

Well, here's a quick first attempt which handles odd/even list sizes. I thought I had a shorter version, but SigmaLIST doesn't handle a list with one element! Ridiculous!

<< SORT DUP SIZE 1 + 2 / DUP FLOOR SWAP PICK3 SWAP GET UNROT GET + 2 / >>

Someone can do a smaller one, can't they?

-Scott

                        
Re: Better Short Median Formula
Message #5 Posted by I, Claudius on 10 Oct 2005, 10:01 a.m.,
in response to message #4 by Scott

Scott: thanks!

      
Re: Stat49Pro+ Impressions Please...
Message #6 Posted by Scott on 10 Oct 2005, 1:27 p.m.,
in response to message #1 by Scott Guth

Thanks to all who have expressed an interest in evaluating this software. The offer is still open for others who may be interested.

Best regards, Scott Guth


[ Return to Index | Top of Index ]

Go back to the main exhibit hall