--
You received this message because you are subscribed to the Google Groups "pystatsmodels" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pystatsmodel...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/pystatsmodels/e9caf5b1-cc71-4fa6-af57-25b8c651307en%40googlegroups.com.
Thanks for your Answer!
And for the hint about categorial variables: Unfortunately I already tried that.
I think it should be one, but declaring it such does not make any difference.I thought I might have missed a default option that differs between the Implementations.
But since I cannot even produce significantly different (wrong) output when changing options, I do not even know where to look.
Paper example:
## Coefficients of interest:
## A B C D delta2 delta3
## 2.071 6.864 2.071 NA 2.390 3.249
My results:
coef std err z P>|z| [0.025 0.975]
----------------------------------------------------------------------------------------
Intercept -32.1278 4627.716 -0.007 0.994 -9102.285 9038.029
C(comparison)[T.2.0] -0.0001 1.683 -6.05e-05 1.000 -3.299 3.298
C(comparison)[T.3.0] -0.4904 1.595 -0.307 0.758 -3.617 2.636
C(comparison)[T.4.0] 0.2562 1.455 0.176 0.860 -2.596 3.108
alt_1 30.9017 4627.716 0.007 0.995 -9039.255 9101.058
alt_2 30.9016 4627.716 0.007 0.995 -9039.255 9101.058
alt_3 29.5082 4627.716 0.006 0.995 -9040.649 9099.666
alt_4 28.2816 4627.717 0.006 0.995 -9041.876 9098.439
delta2 0.5230 1.042 0.502 0.616 -1.519 2.565
delta3 1.0617 46.297 0.023 0.982 -89.678 91.802
========================================================================================
To view this discussion on the web visit https://groups.google.com/d/msgid/pystatsmodels/05a2b9ae-827c-4bbe-a543-7e4210447cfcn%40googlegroups.com.