[R] Help with regression modeling

From: Tony Young <wayoung_at_softbase.math.uwaterloo.ca>
Date: Thu 30 Jun 2005 - 04:51:10 EST


Hello all,  

I'm using R version 2.0.1. I have been having trouble with my linear modeling. I have a table that looks something like this:  

T          RSS     DS        S          LS        PF        COLS   PS
R          RTT      Actual   Max      COMM

char      50000   MSS     2          1250000            1          1
4096     450       0.001    64.30    0.13      64.17

char      50000   MSS     50         1250000            1          25
4096     450       0.001    74.97    0.14      74.83

char      50000   MSS     200       1250000            1          100
4096     450       0.001    111.12  0.13      110.99

char      50000   MSS     3          1250000            1          1
4096     450       0.001    64.31    0.14      64.17

char      50000   MSS     75         1250000            1          25
4096     450       0.001    75.99    0.14      75.85

char      50000   MSS     300       1250000            1          100
4096     450       0.001    116.62  0.14      116.48

 

that I was modeling using the R command  

model.f <- lm(comm ~ I(rtt * ceiling(rss/pf)) + I((s * rss) / ls) + ds*(I(((s+s^2)*rss)/r) + I(((s+s^2)*rss)/(r*ps))))  

R gives me a table like this for coefficients  

              I(rtt * ceiling(rss/pf))                        I((s *
rss)/ls)
                          8.174598e-01

1.353734e+00
                                 dsASA

dsDB2
                          3.101663e+00

2.587372e+00
                                 dsMSS

dsOracle
                          1.575822e+01

9.092041e+00
                I(((s + s^2) * rss)/r)          I(((s + s^2) * rss)/(r *
ps))
                          2.151170e-07

3.087502e-05

          dsDB2:I(((s + s^2) * rss)/r) dsMSS:I(((s + s^2) * rss)/r)

                         -6.972409e-08

1.812149e-07

       dsOracle:I(((s + s^2) * rss)/r) dsDB2:I(((s + s^2) * rss)/(r * ps))

                          1.251699e-08

6.687324e-05

   dsMSS:I(((s + s^2) * rss)/(r * ps)) dsOracle:I(((s + s^2) * rss)/(r * ps))

                          4.731049e-05

-9.243794e-05  

What I want at the end of the day is a formula that says "given this set of parameters, here is the predicted value". When I plug the above coefficients into Excel, I get a value that is WAY off the fitted value that R gives me.  

My questions are:  

  1. How do I use these coefficients in a formula?
  2. How can I construct a formula from these coefficients that I can publish in my thesis?

Thanks for your help!  

ty

        [[alternative HTML version deleted]]



R-help@stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html Received on Thu Jun 30 04:55:25 2005

This archive was generated by hypermail 2.1.8 : Fri 03 Mar 2006 - 03:33:07 EST