I am currently working with non linear regression specifically logistic function. The problem i am facing is that after stating the initial values the outputs remain the same as the default one.
The output remains the same even after specifying b1 as 100.
. nl log3 AllIndia Year, parameters(b0) initial ( b0 40 )
(obs = 12)
The output remains the same even after specifying b1 as 100.
. nl log3 AllIndia Year, parameters(b0) initial ( b0 40 )
(obs = 12)
Code:
Iteration 0: residual SS = 306.8853
Iteration 1: residual SS = 50.04118
Iteration 2: residual SS = 19.38998
Iteration 3: residual SS = 14.59315
Iteration 4: residual SS = 14.56964
Iteration 5: residual SS = 14.5694
Iteration 6: residual SS = 14.5694
Iteration 7: residual SS = 14.5694
Source | SS df MS Number of obs = 12
-------------+------------------------------ F( 3, 9) = 1968.35
Model | 9559.24737 3 3186.41579 Prob > F = 0.0000
Residual | 14.5694011 9 1.61882235 R-squared = 0.9985
-------------+------------------------------ Adj R-squared = 0.9980
Total | 9573.81677 12 797.818064 Root MSE = 1.272329
Res. dev. = 36.38273
3-parameter logistic function, AllIndia=b1/(1+exp(-b2*(Year-b3)))
------------------------------------------------------------------------------
AllIndia | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
b1 | 47.68555 1.191695 40.01 0.000 44.98975 50.38135
b2 | .7198174 .0500143 14.39 0.000 .6066772 .8329576
b3 | 2009.956 .1268755 1.6e+04 0.000 2009.669 2010.243
------------------------------------------------------------------------------
(SEs, P values, CIs, and correlations are asymptotic approximations)