[Limdep Nlogit List] Weighting in quantile regression
William Greene
wgreene at stern.nyu.edu
Sun Mar 6 07:52:04 AEDT 2022
Quantile regression is computed using linear programming, not by any kind
of moments. There is no way to incorporating a "weighting" arrangement.
Weighting request is ignored by QREG. In principle, you could scale the
data.
But, an observation that gets a higher scale still does not obtain greater
prominence in the estimator.
/B. Greene
On Fri, Mar 4, 2022 at 11:39 AM Alessandro Corsi <alessandro.corsi at unito.it>
wrote:
> Does the WTS; specification work with the QREG; model, and with the
> Quantiles; command? (I am using Nlogit 6) The WTS specification seems to
> produce no effect on the estimation.
>
> I run the following: QREG;Lhs=meatprot; Rhs=one,gdppc;quantile=.25 $
> (output below)
>
> and obtained exactly the same output with: QREG;Lhs=meatprot;
> Rhs=one,gdppc;quantile=.25; wts= populati $
>
> (though it signals a weighting variable, see below)
>
> Moreover, when checking the data with: QUANTILES; rhs=meatprot$
>
> I got exactly the same results as with: QUANTILES; rhs=meatprot; wts=
> populati $
>
> Thanks for clarifications
>
> Alessandro Corsi
>
>
> QREG;Lhs=meatprot; Rhs=one,gdppc;quantile=.25 $
>
>
> -----------------------------------------------------------------------------
> Quantile Regression Model. Quantile = .250000
> Linear Programming estimation method
> LHS=MEATPROT Mean = 16.28292
> Standard deviation = 10.97889
> Number of observs. = 2296
> Minimum = 1.16000
> t= .25000 quantile = 6.29000
> Maximum = 46.93000
> Model size Parameters = 2
> Degrees of freedom = 2294
> Residuals Sum of squares = 160762.75237
> Standard error of e = 7.06346
> Fit R-squared = .58608
> PseudoR2=1-F(0)/F(b) = .37028
> Not using OLS or no constant. Rsquared may be <= 0
> Functions F= Sum r(t)[y(i)-x(i)b] = 4459.21969
> F0=Sum r(t)[y(i)-Qy(t)] = 7081.28500
> r(t)[u]=t*u-u*[u<0].t= .250000
> Asymptotic cov. matrix based on kernel estimator.
> Heteroscedasticity test, Chi2[ 1] =237.11 P = .000
>
> --------+--------------------------------------------------------------------
> | Standard Prob. 95% Confidence
> MEATPROT| Coefficient Error z |z|>Z* Interval
>
> --------+--------------------------------------------------------------------
> Constant| 3.27656*** .20597 15.91 .0000 2.87286 3.68026
> GDPPC| .00048*** .9048D-05 53.21 .0000 .00046 .00050
>
> --------+--------------------------------------------------------------------
> nnnnn.D-xx or D+xx => multiply by 10 to -xx or +xx.
> ***, **, * ==> Significance at 1%, 5%, 10% level.
> Model was estimated on Mar 04, 2022 at 04:43:31 PM
>
> -----------------------------------------------------------------------------
>
>
> -----------------------------------------------------------------------------
> Quantile Regression Model. Quantile = .250000
> Linear Programming estimation method
> LHS=MEATPROT Mean = 16.28292
> Standard deviation = 10.97889
> WTS=POPULATI Number of observs. = 2296
> Minimum = 1.16000
> t= .25000 quantile = 6.29000
> Maximum = 46.93000
> Model size Parameters = 2
> Degrees of freedom = 2294
> Residuals Sum of squares = 160762.75237
> Standard error of e = 7.06346
> Fit R-squared = .58608
> PseudoR2=1-F(0)/F(b) = .37028
> Not using OLS or no constant. Rsquared may be <= 0
> Functions F= Sum r(t)[y(i)-x(i)b] = 4459.21969
> F0=Sum r(t)[y(i)-Qy(t)] = 7081.28500
> r(t)[u]=t*u-u*[u<0].t= .250000
> Asymptotic cov. matrix based on kernel estimator.
> Heteroscedasticity test, Chi2[ 1] =237.11 P = .000
>
> --------+--------------------------------------------------------------------
> | Standard Prob. 95% Confidence
> MEATPROT| Coefficient Error z |z|>Z* Interval
>
> --------+--------------------------------------------------------------------
> Constant| 3.27656*** .20597 15.91 .0000 2.87286 3.68026
> GDPPC| .00048*** .9048D-05 53.21 .0000 .00046 .00050
>
> --------+--------------------------------------------------------------------
> nnnnn.D-xx or D+xx => multiply by 10 to -xx or +xx.
> ***, **, * ==> Significance at 1%, 5%, 10% level.
> Model was estimated on Mar 04, 2022 at 04:43:59 PM
>
> -----------------------------------------------------------------------------
>
>
> QREG;Lhs=meatprot; Rhs=one,gdppc;quantile=.25; wts= populati $
>
>
> -----------------------------------------------------------------------------
> Quantile Regression Model. Quantile = .250000
> Linear Programming estimation method
> LHS=MEATPROT Mean = 16.28292
> Standard deviation = 10.97889
> WTS=POPULATI Number of observs. = 2296
> Minimum = 1.16000
> t= .25000 quantile = 6.29000
> Maximum = 46.93000
> Model size Parameters = 2
> Degrees of freedom = 2294
> Residuals Sum of squares = 160762.75237
> Standard error of e = 7.06346
> Fit R-squared = .58608
> PseudoR2=1-F(0)/F(b) = .37028
> Not using OLS or no constant. Rsquared may be <= 0
> Functions F= Sum r(t)[y(i)-x(i)b] = 4459.21969
> F0=Sum r(t)[y(i)-Qy(t)] = 7081.28500
> r(t)[u]=t*u-u*[u<0].t= .250000
> Asymptotic cov. matrix based on kernel estimator.
> Heteroscedasticity test, Chi2[ 1] =237.11 P = .000
>
> --------+--------------------------------------------------------------------
> | Standard Prob. 95% Confidence
> MEATPROT| Coefficient Error z |z|>Z* Interval
>
> --------+--------------------------------------------------------------------
> Constant| 3.27656*** .20597 15.91 .0000 2.87286 3.68026
> GDPPC| .00048*** .9048D-05 53.21 .0000 .00046 .00050
>
> --------+--------------------------------------------------------------------
> nnnnn.D-xx or D+xx => multiply by 10 to -xx or +xx.
> ***, **, * ==> Significance at 1%, 5%, 10% level.
> Model was estimated on Mar 04, 2022 at 04:43:59 PM
>
> -----------------------------------------------------------------------------
>
>
> -----------------------------------------------------------------------------
> Quantile Regression Model. Quantile = .250000
> Linear Programming estimation method
> LHS=MEATPROT Mean = 16.28292
> Standard deviation = 10.97889
> WTS=POPULATI Number of observs. = 2296
> Minimum = 1.16000
> t= .25000 quantile = 6.29000
> Maximum = 46.93000
> Model size Parameters = 2
> Degrees of freedom = 2294
> Residuals Sum of squares = 160762.75237
> Standard error of e = 7.06346
> Fit R-squared = .58608
> PseudoR2=1-F(0)/F(b) = .37028
> Not using OLS or no constant. Rsquared may be <= 0
> Functions F= Sum r(t)[y(i)-x(i)b] = 4459.21969
> F0=Sum r(t)[y(i)-Qy(t)] = 7081.28500
> r(t)[u]=t*u-u*[u<0].t= .250000
> Asymptotic cov. matrix based on kernel estimator.
> Heteroscedasticity test, Chi2[ 1] =237.11 P = .000
>
> --------+--------------------------------------------------------------------
> | Standard Prob. 95% Confidence
> MEATPROT| Coefficient Error z |z|>Z* Interval
>
> --------+--------------------------------------------------------------------
> Constant| 3.27656*** .20597 15.91 .0000 2.87286 3.68026
> GDPPC| .00048*** .9048D-05 53.21 .0000 .00046 .00050
>
> --------+--------------------------------------------------------------------
> nnnnn.D-xx or D+xx => multiply by 10 to -xx or +xx.
> ***, **, * ==> Significance at 1%, 5%, 10% level.
> Model was estimated on Mar 04, 2022 at 04:43:59 PM
>
> -----------------------------------------------------------------------------
>
>
>
> Moreover, when checking the data with: QUANTILES; rhs=meatprot$
>
> I got exactly the same results as with: QUANTILES; rhs=meatprot; wts=
> populati $
>
> Thanks for clarifications
>
> Alessandro Corsi
>
>
> --
> Questa e-mail è stata controllata per individuare virus con Avast
> antivirus.
> https://protect-au.mimecast.com/s/hpssC0YKPviG74YVoFwNXHk?domain=avast.com
> _______________________________________________
> Limdep site list
> Limdep at mailman.sydney.edu.au
> https://protect-au.mimecast.com/s/hLtaCgZ0N1iAoN490ToKcei?domain=limdep.itls.usyd.edu.au
>
--
William Greene
Department of Economics, emeritus
Stern School of Business, New York University
44 West 4 St.
New York, NY, 10012
URL: https://protect-au.mimecast.com/s/fGeqCjZ1N7inBAZ85U5K8VM?domain=people.stern.nyu.edu
Email: wgreene at stern.nyu.edu
Editor in Chief: Journal of Productivity Analysis
<https://protect-au.mimecast.com/s/Aq6-Ck81N9tOK4o9Ws9CuBj?domain=springer.com>
<https://protect-au.mimecast.com/s/Aq6-Ck81N9tOK4o9Ws9CuBj?domain=springer.com>
Editor in Chief: Foundations and Trends in Econometrics
Associate Editor: Economics Letters
Associate Editor: Journal of Business and Economic Statistics
More information about the Limdep
mailing list