st: fraqpoly question

["Salkov, Rick" <Rick.Salkov@eia.doe.gov>]

I have a simple dataset consisting of 2 variables a and b.  I suspect that the fit is not linear so I used fraqpoly. 

The Stata help was a little vague on the generated formula. 

 

 

Here is the run.

 

. fracpoly regress b a

........

-> gen double Ia__1 = X^-2-1.194 if e(sample)

-> gen double Ia__2 = X^3-.7661 if e(sample)

   (where: X = a/100)

 

      Source |       SS       df       MS              Number of obs =      46

-------------+------------------------------           F(  2,    43) =   60.96

       Model |  .011742256     2  .005871128           Prob > F      =  0.0000

    Residual |  .004141086    43  .000096304           R-squared     =  0.7393

-------------+------------------------------           Adj R-squared =  0.7272

       Total |  .015883342    45  .000352963           Root MSE      =  .00981

 

------------------------------------------------------------------------------

           b |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]

-------------+----------------------------------------------------------------

       Ia__1 |  -.0001748   .0004779    -0.37   0.716    -.0011385    .0007889

       Ia__2 |  -.0138219   .0014766    -9.36   0.000    -.0167998   -.0108439

       _cons |   .1115939   .0018996    58.75   0.000      .107763    .1154247

------------------------------------------------------------------------------

Deviance:-297.9678. Best powers of a among 44 models fit: -2 3.

 

After gen x = a/100, I thought that the regression formula would be:

 

 b = -.0001748*a^(x^-2-1.194) - .0138219*a^(x^3-.7661) + .1115939

 

but that formula gives me incorrect results for b.   Can anyone give me the correct formula or suggest any other procs to use?