Forum: help

RE: micEconAids - LA-AIDS model [ Reply ] By: Arne Henningsen on 2019-03-28 08:47	[forum:46634]
Yes, you are right :-)

RE: micEconAids - LA-AIDS model [ Reply ]
By: Arne Henningsen on 2019-03-28 08:46

[forum:46633]

Dear Camilla

The documentation of aidsEst() indicates (among other things) the default values of the arguments, e.g., "hom = TRUE, sym = TRUE", i.e., if you do not specify arguments "'hom' and 'sym', they are both set to 'TRUE', i.e., homogeneity and symmetry are imposed. Hence, as you can see from the regression output on page 5, all homogeneity and symmetry conditions are exactly fulfilled in 'laaidsResult'.

Best regards,
Arne

RE: micEconAids - LA-AIDS model [ Reply ]
By: Camila Brown on 2019-03-27 21:14

[forum:46632]

Dear Arne,

After experimenting with the arguments, I think the answer is YES...when we do not use the arguments hom=TRUE and sym=TRUE, aidsEst() will estimate a CONSTRAINED model since these two arguments are set by default. So, in order to get UNCONSTRAINED estimates we as user sof your package have to set hom=FALSE and sym=FALSE. Otherwise, the coefficients of the AIDS model (linear or nonlinear) will be CONSTRAINED.

Could you please respond with a brief YES or NO to my guess above? This will be helpful to other users.

Sincerely,
Camila

RE: micEconAids - LA-AIDS model [ Reply ]
By: Camila Brown on 2019-03-27 16:18

[forum:46629]

Dear Arne,

Yes, I read that documentation and that one can use arguments "hom" and "sym" to choose whether aidsEst() should impose homogeneity and/or symmetry, but I was not sure if the results that you provide on p. 5 or 6 are unconstrained estimates or constrained estimates since even though you do not specify them on p. 5 in your "laaidsResult" object, I was not sure if the arguments "hom" and "sym" are applied by default or not.

Based on your answer it seems that these coefficients are unconstrained since you did not explicitly use the arguments "hom" and/or "sym" in that object.

Am I right?

Sincerely,
Camila

RE: micEconAids - LA-AIDS model [ Reply ]
By: Arne Henningsen on 2019-03-27 07:44

[forum:46628]

Dear Camila

Please read the documentation of aidsEst(): you can use arguments "hom" and "sym" to choose whether aidsEst() should impose homogeneity and symmetry, respectively.

Currently, there is no newer version of the micEconAids package than the one on CRAN. You can follow (or contribute to) the development of the micEconAids package through R-Forge, e.g.:

https://r-forge.r-project.org/projects/micecon/

https://r-forge.r-project.org/R/?group_id=257

https://r-forge.r-project.org/scm/viewvc.php/pkg/micEconAids/?root=micecon

Best regards,
Arne

micEconAids - LA-AIDS model [ Reply ]
By: Camila Brown on 2019-03-27 02:14

[forum:46627]

Dear Arne,

I am wondering if your illustration of the LA-AIDS model on p. 5 of your document https://cran.r-project.org/web/packages/micEconAids/vignettes/micEconAids_vignette.pdf using:

laaidsResult <- aidsEst( priceNames, shareNames, "xFood", data = Blanciforti86,priceIndex = "S" )

imposes homogeneity and/or symmetry.

1. If so, how can I estimate the same model without any constraints? So, I get unconstrained estimates.
2. If not, how can impose homogeneity, symmetry or both?

Sincerely,
Camila

BTW, is there a more recent https://cran.r-project.org/web/packages/micEconAids/micEconAids.pdf This one is from 2017