From kchakraborty at uwf.edu Sun Jul 9 00:27:46 2017 From: kchakraborty at uwf.edu (Kalyan Chakraborty) Date: Sat, 8 Jul 2017 09:27:46 -0500 Subject: [Limdep Nlogit List] Hausman Taylor Est Message-ID: Hi Everyone, I am trying to estimate H-T Estimator (page E11-79 Version-9) with 30 country, 12 year, 11 variable a shorter version than my original data that has 93 countries, 22 years and 23 variables. It did well until I reached 2SLS results. I am getting an error message,Error-608 which according to the manual is ---AR(1) model cannot be computed with weighting. Also, can I use lagged dependent variable as: Namelist; F1=one,Y[-1] with no F2. Any help would be appreciated. Thank you, Kal |-> CALC; S2S=SSQRD; S2U=S2S-AVG_TI*S2E$ |-> REGRESS; LHS=Y; RHS=ALL; PANEL; RANDOM; PDS=12; START=KX1,KX2,KF1,S2E,S2U$ Error 608: No room in data area to create needed group indicator *Dr. Kalyan (Kal) Chakraborty* From kchakraborty at uwf.edu Tue Jul 11 02:54:56 2017 From: kchakraborty at uwf.edu (Kalyan Chakraborty) Date: Mon, 10 Jul 2017 11:54:56 -0500 Subject: [Limdep Nlogit List] Need help Message-ID: Dear Prof. Green and others, I posted a message for help on the error message 2 days ago, I got from running the Hausman-Taylor estimation (V9 page E11-79). But I am using NLOGIT5 on a small set of data. The code is reproduced below. I also expanded the data area using Tools-Project-Memory Allocation-Data Cells to 5000,000 but is is still giving the error message 608. It ran well but stopped as soon as it reached the last two lines. Any help will be appreciated. Kal +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ OPEN ; OUTPUT = F:\NEWLIM\SHORT.OUT$ READ ; FILE = F:\NEWLIM\SHORT.TXT ; NVAR = 11; NOBS = 360 ; NAMES = YEAR,COUNTRY,GROW,GNI,LID,NLID,INFLA,GEXP,OPEN,TENROL,TRADE$ DSTAT; RHS = GROW,GNI,LID,NLID,INFLA,GEXP,OPEN,TENROL,TRADE$ CREATE; LGNI=LOG(GNI)$ CREATE; LGROW=GROW[-1]$ NAMELIST; X1=LGROW,LGNI,GEXP$ NAMELIST; X2=OPEN,TENROL,TRADE$ NAMELIST; F1=ONE,NLID,LID$ NAMELIST; F2=INFLA$ NAMELIST; X=X1,X2;F=F1,F2; EXOG=X1,F1;ALL=X1,X2,F1,F2$ CREATE; Y = GROW$ CALC; KX1=COL(X1); KX2=COL(X2);KF1=COL(F1);KF2=COL(F2)$ REGRESS; LHS=Y; RHS=X; PDS=12; FIXED;PANEL$ MATRIX; BW=B$ CALC; S2E=SSQRD$ CREATE; DWIT = Y-X'BW$ REGRESS; LHS=DWIT; RHS=ONE; PDS =12; PANEL; KEEP=DWI$ 2SLS; LHS=DWI; RHS=F; INST=EXOG$ CALC; S2S=SSQRD; S2U=S2S-AVG_TI*S2E$ REGRESS; LHS=Y; RHS=ALL; PANEL; RANDOM; PDS=12; START=KX1,KX2,KF1,S2E,S2U$ |-> CALC; S2S=SSQRD; S2U=S2S-AVG_TI*S2E$ |-> REGRESS; LHS=Y; RHS=ALL; PANEL; RANDOM; PDS=12; START=KX1,KX2,KF1,S2E,S2U$ Error 608: No room in data area to create needed group indicator ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ *Dr. Kalyan (Kal) Chakraborty* From richard.turner at imarketresearch.com Wed Jul 12 04:04:05 2017 From: richard.turner at imarketresearch.com (Richard Turner) Date: Tue, 11 Jul 2017 14:04:05 -0400 Subject: [Limdep Nlogit List] How are attribute nonattendence models estimated in NLOGIT? Message-ID: Greetings, I have fielded a choice study and I have information on which attributes respondents explicitly ignored when making choices. After doing some research, I found that I could code each attribute that was explicitly ignored as -888 in NLOGIT that induces NLOGIT to switch to a "attribute nonattendence model." To my knowledge the attribute nonattendence model explicitly fixes coefficients at the individual level to 0. E.g. if the utility function was: U_ij = b_1*COLOR_ij + b_2*Price_ij and the individual ignored color, then the utility function would become U_ij = 0*COLOR_ij + b_2*Price_ij Therefore, the sum of the utility that would enter the likelihood function would only consist of b_2*Price_ij. My questions are: 1. It seems that modeling these nonattended attributes using the method above is equivalent to modeling these attributes as *missing at random*. I thought that if data that was *missing at random* was modeled in this way, it would produce biased estimates. How can this method then not produce inconsistent and biased parameter estimates? 2. Is it possible to estimate *mixed-logit attribute nonattendence models* in NLOGIT? 3. Does anyone know of resources that give more of an explanation of these models, e.g. why they would produce consistent and unbiased parameter estimates? ______________________________________________ *Richard Turner* *iMarketResearch* *Easier, Faster, Better Advanced Market Research* www.imarketresearch.com Address: 12 Penns Trail, Suite A, Newtown, PA 18940 From trentspears at hotmail.com Mon Jul 17 12:06:21 2017 From: trentspears at hotmail.com (Trent Spears) Date: Mon, 17 Jul 2017 02:06:21 +0000 Subject: [Limdep Nlogit List] Model selection Message-ID: Hi, Is it possible to estimate a (multinomial logistic) regression model in LIMDEP using a process of backward elimination? In particular, I would like to specify a threshold p-value and retrieve model output including only the independent variables with a p-value below this threshold. For example, the sample code below might return the top-most table. What do I add to get the bottom-most table instead? Thank you in advance! Sample code, for p-value threshold = 0.10: DISCRETECHOICE;Lhs=Choice,Group;Conditional; Rhs = var1,var2,var3$ --------+-------------------------------------------------------------------- | Standard Prob. 95% Confidence PLACE| Coefficient Error z |z|>Z* Interval --------+-------------------------------------------------------------------- VAR1| -3.96840* 2.38165 -1.67 .0957 -8.63634 .69954 VAR2| .40056 .69390 .58 .5638 -.95947 1.76058 VAR3| -.02381 1.10233 -.02 .9828 -2.18433 2.13672 --------+-------------------------------------------------------------------- | Standard Prob. 95% Confidence PLACE| Coefficient Error z |z|>Z* Interval --------+-------------------------------------------------------------------- VAR1| V* W X .082 Y Z From wgreene at stern.nyu.edu Mon Jul 17 20:06:43 2017 From: wgreene at stern.nyu.edu (William Greene) Date: Mon, 17 Jul 2017 06:06:43 -0400 Subject: [Limdep Nlogit List] Model selection In-Reply-To: References: Message-ID: Trent: In a multinomial choice model, you would not typically think of the attributes as if they were regressors in a conditional mean function. As such, this seems like a counterproductive way to search for the functional form of the utility functions, assuming the list of attributes that was specified at the beginning was considered appropriate at the start. There is a literature on "attribute non-attendance" that might be helpful. All that said, I'm afraid not. NLOGIT (LIMDEP) does not provide an automated backward stepwise procedure for finding the specification of the model. Regards Bill Greene On Sun, Jul 16, 2017 at 10:06 PM, Trent Spears wrote: > Hi, > > > Is it possible to estimate a (multinomial logistic) regression model in > LIMDEP using a process of backward elimination? In particular, I would like > to specify a threshold p-value and retrieve model output including only the > independent variables with a p-value below this threshold. > > For example, the sample code below might return the top-most table. What > do I add to get the bottom-most table instead? > > > Thank you in advance! > > > > Sample code, for p-value threshold = 0.10: > > DISCRETECHOICE;Lhs=Choice,Group;Conditional; Rhs = var1,var2,var3$ > > > --------+--------------------------------------------------- > ----------------- > | Standard Prob. 95% Confidence > PLACE| Coefficient Error z |z|>Z* Interval > --------+--------------------------------------------------- > ----------------- > VAR1| -3.96840* 2.38165 -1.67 .0957 -8.63634 .69954 > VAR2| .40056 .69390 .58 .5638 -.95947 1.76058 > VAR3| -.02381 1.10233 -.02 .9828 -2.18433 2.13672 > > > --------+--------------------------------------------------- > ----------------- > | Standard Prob. 95% Confidence > PLACE| Coefficient Error z |z|>Z* Interval > --------+--------------------------------------------------- > ----------------- > VAR1| V* W X .082 Y Z > > > > _______________________________________________ > Limdep site list > Limdep at limdep.itls.usyd.edu.au > http://limdep.itls.usyd.edu.au > > -- William Greene Department of Economics Stern School of Business, New York University 44 West 4 St., 7-90 New York, NY, 10012 URL: http://people.stern.nyu.edu/wgreene Email: wgreene at stern.nyu.edu Ph. +1.212.998.0876 Editor in Chief: Journal of Productivity Analysis Editor in Chief: Foundations and Trends in Econometrics Associate Editor: Economics Letters Associate Editor: Journal of Business and Economic Statistics Associate Editor: Journal of Choice Modeling From wgreene at stern.nyu.edu Mon Jul 17 20:46:13 2017 From: wgreene at stern.nyu.edu (William Greene) Date: Mon, 17 Jul 2017 06:46:13 -0400 Subject: [Limdep Nlogit List] Need help In-Reply-To: References: Message-ID: Dear Kal. That diagnostic looks to me to be erroneous. I suspect the real error is that you did not provide KF2 in your regress command. /Bill Greene On Mon, Jul 10, 2017 at 12:54 PM, Kalyan Chakraborty wrote: > Dear Prof. Green and others, > > I posted a message for help on the error message 2 days ago, I got from > running the Hausman-Taylor estimation (V9 page E11-79). But I am using > NLOGIT5 on a small set of data. The code is reproduced below. I also > expanded the data area using Tools-Project-Memory Allocation-Data Cells to > 5000,000 but is is still giving the error message 608. It ran well but > stopped as soon as it reached the last two lines. > > Any help will be appreciated. > > Kal > > +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ > > OPEN ; OUTPUT = F:\NEWLIM\SHORT.OUT$ > > READ ; FILE = F:\NEWLIM\SHORT.TXT > > ; NVAR = 11; NOBS = 360 > > ; NAMES = > YEAR,COUNTRY,GROW,GNI,LID,NLID,INFLA,GEXP,OPEN,TENROL,TRADE$ > > DSTAT; RHS = GROW,GNI,LID,NLID,INFLA,GEXP,OPEN,TENROL,TRADE$ > > CREATE; LGNI=LOG(GNI)$ > > CREATE; LGROW=GROW[-1]$ > > NAMELIST; X1=LGROW,LGNI,GEXP$ > > NAMELIST; X2=OPEN,TENROL,TRADE$ > > NAMELIST; F1=ONE,NLID,LID$ > > NAMELIST; F2=INFLA$ > > NAMELIST; X=X1,X2;F=F1,F2; EXOG=X1,F1;ALL=X1,X2,F1,F2$ > > CREATE; Y = GROW$ > > CALC; KX1=COL(X1); KX2=COL(X2);KF1=COL(F1);KF2=COL(F2)$ > > REGRESS; LHS=Y; RHS=X; PDS=12; FIXED;PANEL$ > > MATRIX; BW=B$ > > CALC; S2E=SSQRD$ > > CREATE; DWIT = Y-X'BW$ > > REGRESS; LHS=DWIT; RHS=ONE; PDS =12; PANEL; KEEP=DWI$ > > 2SLS; LHS=DWI; RHS=F; INST=EXOG$ > > CALC; S2S=SSQRD; S2U=S2S-AVG_TI*S2E$ > > REGRESS; LHS=Y; RHS=ALL; PANEL; RANDOM; PDS=12; START=KX1,KX2,KF1,S2E,S2U$ > > > |-> CALC; S2S=SSQRD; S2U=S2S-AVG_TI*S2E$ > > |-> REGRESS; LHS=Y; RHS=ALL; PANEL; RANDOM; PDS=12; > START=KX1,KX2,KF1,S2E,S2U$ > > Error 608: No room in data area to create needed group indicator > > > ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ > > *Dr. Kalyan (Kal) Chakraborty* > _______________________________________________ > Limdep site list > Limdep at limdep.itls.usyd.edu.au > http://limdep.itls.usyd.edu.au > > -- William Greene Department of Economics Stern School of Business, New York University 44 West 4 St., 7-90 New York, NY, 10012 URL: http://people.stern.nyu.edu/wgreene Email: wgreene at stern.nyu.edu Ph. +1.212.998.0876 Editor in Chief: Journal of Productivity Analysis Editor in Chief: Foundations and Trends in Econometrics Associate Editor: Economics Letters Associate Editor: Journal of Business and Economic Statistics Associate Editor: Journal of Choice Modeling From kchakraborty at uwf.edu Mon Jul 17 22:30:52 2017 From: kchakraborty at uwf.edu (Kalyan Chakraborty) Date: Mon, 17 Jul 2017 07:30:52 -0500 Subject: [Limdep Nlogit List] Need help In-Reply-To: References: Message-ID: Prof. Greene, Thank you for your response. I was looking at the manual (page E11-86) says if I do not have variables for all 4 sets of functions such as no time invariant variable then I can set KF1=0... can I set KF2=0 too? Thank you again, Kal *Dr. Kalyan (Kal) Chakraborty* *Associate Director, Haas Center* *Business Research and Economic Development* *University of West Florida* *220 West Garden St,Suite 303* *Pensacola, FL 32502* *(P) 850- 439-5418 (E) kchakraborty at uwf.edu * https://secure.uwf.edu/cob/ http://haas.uwf.edu/Home/Staff#chakraborty On Mon, Jul 17, 2017 at 5:46 AM, William Greene wrote: > Dear Kal. > That diagnostic looks to me to be erroneous. I suspect the real error is > that you did not > provide KF2 in your regress command. > /Bill Greene > > On Mon, Jul 10, 2017 at 12:54 PM, Kalyan Chakraborty > > wrote: > > > Dear Prof. Green and others, > > > > I posted a message for help on the error message 2 days ago, I got from > > running the Hausman-Taylor estimation (V9 page E11-79). But I am using > > NLOGIT5 on a small set of data. The code is reproduced below. I also > > expanded the data area using Tools-Project-Memory Allocation-Data Cells > to > > 5000,000 but is is still giving the error message 608. It ran well but > > stopped as soon as it reached the last two lines. > > > > Any help will be appreciated. > > > > Kal > > > > +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ > > > > OPEN ; OUTPUT = F:\NEWLIM\SHORT.OUT$ > > > > READ ; FILE = F:\NEWLIM\SHORT.TXT > > > > ; NVAR = 11; NOBS = 360 > > > > ; NAMES = > > YEAR,COUNTRY,GROW,GNI,LID,NLID,INFLA,GEXP,OPEN,TENROL,TRADE$ > > > > DSTAT; RHS = GROW,GNI,LID,NLID,INFLA,GEXP,OPEN,TENROL,TRADE$ > > > > CREATE; LGNI=LOG(GNI)$ > > > > CREATE; LGROW=GROW[-1]$ > > > > NAMELIST; X1=LGROW,LGNI,GEXP$ > > > > NAMELIST; X2=OPEN,TENROL,TRADE$ > > > > NAMELIST; F1=ONE,NLID,LID$ > > > > NAMELIST; F2=INFLA$ > > > > NAMELIST; X=X1,X2;F=F1,F2; EXOG=X1,F1;ALL=X1,X2,F1,F2$ > > > > CREATE; Y = GROW$ > > > > CALC; KX1=COL(X1); KX2=COL(X2);KF1=COL(F1);KF2=COL(F2)$ > > > > REGRESS; LHS=Y; RHS=X; PDS=12; FIXED;PANEL$ > > > > MATRIX; BW=B$ > > > > CALC; S2E=SSQRD$ > > > > CREATE; DWIT = Y-X'BW$ > > > > REGRESS; LHS=DWIT; RHS=ONE; PDS =12; PANEL; KEEP=DWI$ > > > > 2SLS; LHS=DWI; RHS=F; INST=EXOG$ > > > > CALC; S2S=SSQRD; S2U=S2S-AVG_TI*S2E$ > > > > REGRESS; LHS=Y; RHS=ALL; PANEL; RANDOM; PDS=12; > START=KX1,KX2,KF1,S2E,S2U$ > > > > > > |-> CALC; S2S=SSQRD; S2U=S2S-AVG_TI*S2E$ > > > > |-> REGRESS; LHS=Y; RHS=ALL; PANEL; RANDOM; PDS=12; > > START=KX1,KX2,KF1,S2E,S2U$ > > > > Error 608: No room in data area to create needed group indicator > > > > > > ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ > > > > *Dr. Kalyan (Kal) Chakraborty* > > _______________________________________________ > > Limdep site list > > Limdep at limdep.itls.usyd.edu.au > > http://limdep.itls.usyd.edu.au > > > > > > > -- > William Greene > Department of Economics > Stern School of Business, New York University > 44 West 4 St., 7-90 > New York, NY, 10012 > URL: http://people.stern.nyu.edu/wgreene > Email: wgreene at stern.nyu.edu > Ph. +1.212.998.0876 > Editor in Chief: Journal of Productivity Analysis > Editor in Chief: Foundations and Trends in Econometrics > Associate Editor: Economics Letters > Associate Editor: Journal of Business and Economic Statistics > Associate Editor: Journal of Choice Modeling > _______________________________________________ > Limdep site list > Limdep at limdep.itls.usyd.edu.au > http://limdep.itls.usyd.edu.au > > From wgreene at stern.nyu.edu Tue Jul 18 08:15:25 2017 From: wgreene at stern.nyu.edu (William Greene) Date: Mon, 17 Jul 2017 18:15:25 -0400 Subject: [Limdep Nlogit List] Need help In-Reply-To: References: Message-ID: Kal, it can be zero. But, you did not specify it at all. You must include all four values. .B. Greene On Mon, Jul 17, 2017 at 8:30 AM, Kalyan Chakraborty wrote: > Prof. Greene, > > Thank you for your response. I was looking at the manual (page E11-86) says > if I do not have variables for all 4 sets of functions such as no time > invariant variable then I can set KF1=0... can I set KF2=0 too? > > Thank you again, > Kal > > > > *Dr. Kalyan (Kal) Chakraborty* > > *Associate Director, Haas Center* > > *Business Research and Economic Development* > > *University of West Florida* > > *220 West Garden St,Suite 303* > > *Pensacola, FL 32502* > > *(P) 850- 439-5418 (E) kchakraborty at uwf.edu * > https://secure.uwf.edu/cob/ > http://haas.uwf.edu/Home/Staff#chakraborty > > > > > On Mon, Jul 17, 2017 at 5:46 AM, William Greene > wrote: > > > Dear Kal. > > That diagnostic looks to me to be erroneous. I suspect the real error is > > that you did not > > provide KF2 in your regress command. > > /Bill Greene > > > > On Mon, Jul 10, 2017 at 12:54 PM, Kalyan Chakraborty < > kchakraborty at uwf.edu > > > > > wrote: > > > > > Dear Prof. Green and others, > > > > > > I posted a message for help on the error message 2 days ago, I got from > > > running the Hausman-Taylor estimation (V9 page E11-79). But I am using > > > NLOGIT5 on a small set of data. The code is reproduced below. I also > > > expanded the data area using Tools-Project-Memory Allocation-Data Cells > > to > > > 5000,000 but is is still giving the error message 608. It ran well but > > > stopped as soon as it reached the last two lines. > > > > > > Any help will be appreciated. > > > > > > Kal > > > > > > +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ > > > > > > OPEN ; OUTPUT = F:\NEWLIM\SHORT.OUT$ > > > > > > READ ; FILE = F:\NEWLIM\SHORT.TXT > > > > > > ; NVAR = 11; NOBS = 360 > > > > > > ; NAMES = > > > YEAR,COUNTRY,GROW,GNI,LID,NLID,INFLA,GEXP,OPEN,TENROL,TRADE$ > > > > > > DSTAT; RHS = GROW,GNI,LID,NLID,INFLA,GEXP,OPEN,TENROL,TRADE$ > > > > > > CREATE; LGNI=LOG(GNI)$ > > > > > > CREATE; LGROW=GROW[-1]$ > > > > > > NAMELIST; X1=LGROW,LGNI,GEXP$ > > > > > > NAMELIST; X2=OPEN,TENROL,TRADE$ > > > > > > NAMELIST; F1=ONE,NLID,LID$ > > > > > > NAMELIST; F2=INFLA$ > > > > > > NAMELIST; X=X1,X2;F=F1,F2; EXOG=X1,F1;ALL=X1,X2,F1,F2$ > > > > > > CREATE; Y = GROW$ > > > > > > CALC; KX1=COL(X1); KX2=COL(X2);KF1=COL(F1);KF2=COL(F2)$ > > > > > > REGRESS; LHS=Y; RHS=X; PDS=12; FIXED;PANEL$ > > > > > > MATRIX; BW=B$ > > > > > > CALC; S2E=SSQRD$ > > > > > > CREATE; DWIT = Y-X'BW$ > > > > > > REGRESS; LHS=DWIT; RHS=ONE; PDS =12; PANEL; KEEP=DWI$ > > > > > > 2SLS; LHS=DWI; RHS=F; INST=EXOG$ > > > > > > CALC; S2S=SSQRD; S2U=S2S-AVG_TI*S2E$ > > > > > > REGRESS; LHS=Y; RHS=ALL; PANEL; RANDOM; PDS=12; > > START=KX1,KX2,KF1,S2E,S2U$ > > > > > > > > > |-> CALC; S2S=SSQRD; S2U=S2S-AVG_TI*S2E$ > > > > > > |-> REGRESS; LHS=Y; RHS=ALL; PANEL; RANDOM; PDS=12; > > > START=KX1,KX2,KF1,S2E,S2U$ > > > > > > Error 608: No room in data area to create needed group indicator > > > > > > > > > ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ > > > > > > *Dr. Kalyan (Kal) Chakraborty* > > > _______________________________________________ > > > Limdep site list > > > Limdep at limdep.itls.usyd.edu.au > > > http://limdep.itls.usyd.edu.au > > > > > > > > > > > > -- > > William Greene > > Department of Economics > > Stern School of Business, New York University > > 44 West 4 St., 7-90 > > New York, NY, 10012 > > URL: http://people.stern.nyu.edu/wgreene > > Email: wgreene at stern.nyu.edu > > Ph. +1.212.998.0876 > > Editor in Chief: Journal of Productivity Analysis > > Editor in Chief: Foundations and Trends in Econometrics > > Associate Editor: Economics Letters > > Associate Editor: Journal of Business and Economic Statistics > > Associate Editor: Journal of Choice Modeling > > _______________________________________________ > > Limdep site list > > Limdep at limdep.itls.usyd.edu.au > > http://limdep.itls.usyd.edu.au > > > > > _______________________________________________ > Limdep site list > Limdep at limdep.itls.usyd.edu.au > http://limdep.itls.usyd.edu.au > > -- William Greene Department of Economics Stern School of Business, New York University 44 West 4 St., 7-90 New York, NY, 10012 URL: http://people.stern.nyu.edu/wgreene Email: wgreene at stern.nyu.edu Ph. +1.212.998.0876 Editor in Chief: Journal of Productivity Analysis Editor in Chief: Foundations and Trends in Econometrics Associate Editor: Economics Letters Associate Editor: Journal of Business and Economic Statistics Associate Editor: Journal of Choice Modeling From kchakraborty at uwf.edu Tue Jul 18 13:07:36 2017 From: kchakraborty at uwf.edu (Kalyan Chakraborty) Date: Mon, 17 Jul 2017 22:07:36 -0500 Subject: [Limdep Nlogit List] Need help In-Reply-To: References: Message-ID: Thank you so much now I got your point. I will correct it and then rerun the program. You have a nice day, Kal *Dr. Kalyan (Kal) Chakraborty* *Associate Director, Haas Center* *Business Research and Economic Development* *University of West Florida* *220 West Garden St,Suite 303* *Pensacola, FL 32502* *(P) 850- 439-5418 (E) kchakraborty at uwf.edu * https://secure.uwf.edu/cob/ http://haas.uwf.edu/Home/Staff#chakraborty On Mon, Jul 17, 2017 at 5:15 PM, William Greene wrote: > Kal, it can be zero. But, you did not specify it at all. You must include > all four values. > .B. Greene > > On Mon, Jul 17, 2017 at 8:30 AM, Kalyan Chakraborty > wrote: > > > Prof. Greene, > > > > Thank you for your response. I was looking at the manual (page E11-86) > says > > if I do not have variables for all 4 sets of functions such as no time > > invariant variable then I can set KF1=0... can I set KF2=0 too? > > > > Thank you again, > > Kal > > > > > > > > *Dr. Kalyan (Kal) Chakraborty* > > > > *Associate Director, Haas Center* > > > > *Business Research and Economic Development* > > > > *University of West Florida* > > > > *220 West Garden St,Suite 303* > > > > *Pensacola, FL 32502* > > > > *(P) 850- 439-5418 (E) kchakraborty at uwf.edu * > > https://secure.uwf.edu/cob/ > > http://haas.uwf.edu/Home/Staff#chakraborty > > > > > > > > > > On Mon, Jul 17, 2017 at 5:46 AM, William Greene > > wrote: > > > > > Dear Kal. > > > That diagnostic looks to me to be erroneous. I suspect the real error > is > > > that you did not > > > provide KF2 in your regress command. > > > /Bill Greene > > > > > > On Mon, Jul 10, 2017 at 12:54 PM, Kalyan Chakraborty < > > kchakraborty at uwf.edu > > > > > > > wrote: > > > > > > > Dear Prof. Green and others, > > > > > > > > I posted a message for help on the error message 2 days ago, I got > from > > > > running the Hausman-Taylor estimation (V9 page E11-79). But I am > using > > > > NLOGIT5 on a small set of data. The code is reproduced below. I also > > > > expanded the data area using Tools-Project-Memory Allocation-Data > Cells > > > to > > > > 5000,000 but is is still giving the error message 608. It ran well > but > > > > stopped as soon as it reached the last two lines. > > > > > > > > Any help will be appreciated. > > > > > > > > Kal > > > > > > > > +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ > > > > > > > > OPEN ; OUTPUT = F:\NEWLIM\SHORT.OUT$ > > > > > > > > READ ; FILE = F:\NEWLIM\SHORT.TXT > > > > > > > > ; NVAR = 11; NOBS = 360 > > > > > > > > ; NAMES = > > > > YEAR,COUNTRY,GROW,GNI,LID,NLID,INFLA,GEXP,OPEN,TENROL,TRADE$ > > > > > > > > DSTAT; RHS = GROW,GNI,LID,NLID,INFLA,GEXP,OPEN,TENROL,TRADE$ > > > > > > > > CREATE; LGNI=LOG(GNI)$ > > > > > > > > CREATE; LGROW=GROW[-1]$ > > > > > > > > NAMELIST; X1=LGROW,LGNI,GEXP$ > > > > > > > > NAMELIST; X2=OPEN,TENROL,TRADE$ > > > > > > > > NAMELIST; F1=ONE,NLID,LID$ > > > > > > > > NAMELIST; F2=INFLA$ > > > > > > > > NAMELIST; X=X1,X2;F=F1,F2; EXOG=X1,F1;ALL=X1,X2,F1,F2$ > > > > > > > > CREATE; Y = GROW$ > > > > > > > > CALC; KX1=COL(X1); KX2=COL(X2);KF1=COL(F1);KF2=COL(F2)$ > > > > > > > > REGRESS; LHS=Y; RHS=X; PDS=12; FIXED;PANEL$ > > > > > > > > MATRIX; BW=B$ > > > > > > > > CALC; S2E=SSQRD$ > > > > > > > > CREATE; DWIT = Y-X'BW$ > > > > > > > > REGRESS; LHS=DWIT; RHS=ONE; PDS =12; PANEL; KEEP=DWI$ > > > > > > > > 2SLS; LHS=DWI; RHS=F; INST=EXOG$ > > > > > > > > CALC; S2S=SSQRD; S2U=S2S-AVG_TI*S2E$ > > > > > > > > REGRESS; LHS=Y; RHS=ALL; PANEL; RANDOM; PDS=12; > > > START=KX1,KX2,KF1,S2E,S2U$ > > > > > > > > > > > > |-> CALC; S2S=SSQRD; S2U=S2S-AVG_TI*S2E$ > > > > > > > > |-> REGRESS; LHS=Y; RHS=ALL; PANEL; RANDOM; PDS=12; > > > > START=KX1,KX2,KF1,S2E,S2U$ > > > > > > > > Error 608: No room in data area to create needed group indicator > > > > > > > > > > > > ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ > > > > > > > > *Dr. Kalyan (Kal) Chakraborty* > > > > _______________________________________________ > > > > Limdep site list > > > > Limdep at limdep.itls.usyd.edu.au > > > > http://limdep.itls.usyd.edu.au > > > > > > > > > > > > > > > > > -- > > > William Greene > > > Department of Economics > > > Stern School of Business, New York University > > > 44 West 4 St., 7-90 > > > New York, NY, 10012 > > > URL: http://people.stern.nyu.edu/wgreene > > > Email: wgreene at stern.nyu.edu > > > Ph. +1.212.998.0876 > > > Editor in Chief: Journal of Productivity Analysis > > > Editor in Chief: Foundations and Trends in Econometrics > > > Associate Editor: Economics Letters > > > Associate Editor: Journal of Business and Economic Statistics > > > Associate Editor: Journal of Choice Modeling > > > _______________________________________________ > > > Limdep site list > > > Limdep at limdep.itls.usyd.edu.au > > > http://limdep.itls.usyd.edu.au > > > > > > > > _______________________________________________ > > Limdep site list > > Limdep at limdep.itls.usyd.edu.au > > http://limdep.itls.usyd.edu.au > > > > > > > -- > William Greene > Department of Economics > Stern School of Business, New York University > 44 West 4 St., 7-90 > New York, NY, 10012 > URL: http://people.stern.nyu.edu/wgreene > Email: wgreene at stern.nyu.edu > Ph. +1.212.998.0876 > Editor in Chief: Journal of Productivity Analysis > Editor in Chief: Foundations and Trends in Econometrics > Associate Editor: Economics Letters > Associate Editor: Journal of Business and Economic Statistics > Associate Editor: Journal of Choice Modeling > _______________________________________________ > Limdep site list > Limdep at limdep.itls.usyd.edu.au > http://limdep.itls.usyd.edu.au > > From Jane.Kolodinsky at uvm.edu Fri Jul 21 00:51:27 2017 From: Jane.Kolodinsky at uvm.edu (Jane Kolodinsky) Date: Thu, 20 Jul 2017 14:51:27 +0000 Subject: [Limdep Nlogit List] Question about bivariate probit Message-ID: <3a2fc4196a2c4dc39ce2907f5d0a0c22@uvm.edu> I am trying to determine whether an individual's opposition to genetic engineering is jointly determined with seeing a label that says "produced using GE." I ran the wald test to "automate" the test and they are contradictory. RHO(1,2)| -.01910 .06371 -.30 .7644 -.14396 .10577 Above is the relevant output from the bivariate probit: bivariateprobit; Lhs=z1, z2; Rh1=x1; rh2= x2$ The prob value is .76---do not reject rho>0. I also calculated the wald test following p. E21-9 in the Limdep manual. calc;list; waldtest=rho^/varrho; pvalue=1-chi(waldtest,1)$ That output is: CALC] WALDTEST= 246.3811130 [CALC] PVALUE = .0000000 Calculator: Computed 2 scalar results The output is telling me not to reject they null hypothesis that rho =0 and the Wald test is telling me to reject the null hypothesis. Please advise. From wgreene at stern.nyu.edu Fri Jul 21 06:35:59 2017 From: wgreene at stern.nyu.edu (William Greene) Date: Thu, 20 Jul 2017 16:35:59 -0400 Subject: [Limdep Nlogit List] Question about bivariate probit In-Reply-To: <3a2fc4196a2c4dc39ce2907f5d0a0c22@uvm.edu> References: <3a2fc4196a2c4dc39ce2907f5d0a0c22@uvm.edu> Message-ID: Jane. (1) RHO is not the correlation of z1 and z2. It is the correlation of the disturbances. Your manual tells you how to compute a tetrachoric correlation. You should use that. (2) Your WALD command is incorrect. The function you want to analyze is RHO over the square root of the variance. Your use of the ^ in the function asks the program to raise rho to the power 1/var(rho). You should just trust the probit output, It is giving you the right test statistic, although, you are not interpreting RHO correctly. /Bill Greene On Thu, Jul 20, 2017 at 10:51 AM, Jane Kolodinsky wrote: > I am trying to determine whether an individual's opposition to genetic > engineering is jointly determined with seeing a label that says "produced > using GE." > > I ran the wald test to "automate" the test and they are contradictory. > > RHO(1,2)| -.01910 .06371 -.30 .7644 -.14396 .10577 > > Above is the relevant output from the bivariate probit: bivariateprobit; > Lhs=z1, z2; Rh1=x1; rh2= x2$ > > The prob value is .76---do not reject rho>0. > > I also calculated the wald test following p. E21-9 in the Limdep manual. > > > calc;list; waldtest=rho^/varrho; pvalue=1-chi(waldtest,1)$ > > > That output is: > > CALC] WALDTEST= 246.3811130 > [CALC] PVALUE = .0000000 > Calculator: Computed 2 scalar results > > The output is telling me not to reject they null hypothesis that rho =0 > and the Wald test is telling me to reject the null hypothesis. > > Please advise. > > > > > _______________________________________________ > Limdep site list > Limdep at limdep.itls.usyd.edu.au > http://limdep.itls.usyd.edu.au > > -- William Greene Department of Economics Stern School of Business, New York University 44 West 4 St., 7-90 New York, NY, 10012 URL: http://people.stern.nyu.edu/wgreene Email: wgreene at stern.nyu.edu Ph. +1.212.998.0876 Editor in Chief: Journal of Productivity Analysis Editor in Chief: Foundations and Trends in Econometrics Associate Editor: Economics Letters Associate Editor: Journal of Business and Economic Statistics Associate Editor: Journal of Choice Modeling From SXB1022 at student.bham.ac.uk Fri Jul 21 20:09:54 2017 From: SXB1022 at student.bham.ac.uk (Saul Basurto Hernandez) Date: Fri, 21 Jul 2017 10:09:54 +0000 Subject: [Limdep Nlogit List] Nested Logit Sequential (Two step) estimation Message-ID: <09629F44AB19A249A1AFEDC0EE0614DCF1387836@EX12.adf.bham.ac.uk> I am trying to estimate a Nested Logit model using the Sequential estimation as the number of alternatives in my database make the FIML estimation not feasible. In order to be familiarized with the software, I tried to fit the same models as in the NLogit 6 Reference Guide (pages N-528 and N-529) but, the results I got from the second step considerably differed with respect to those in the Reference Guide. This is the example in the Reference Guide (using the clogit.lpj project): NLOGIT ; Lhs = mode ; Choices = air,train,bus,car ; Tree = fly(air), ground(train,bus,car) ; Model: U(air,train,bus,car) = <0,at,ab,0> + bc * gc / U(fly,ground) = ah * hinca ; Ivb = incvlu ; Conditional $ I got exactly the same results here but, not for the second step: NLOGIT ; Lhs = mode ; Choices = air,train,bus,car ; Tree = fly(air), ground(train,bus,car) ; Model: U(air,train,bus,car) = <0,at,ab,0> + bc * gc / U(fly,ground) = ah * hinca + aiv * incvlu ; Sequential $ This is what the reference displays: ----------------------------------------------------------------------------- Second step estimates of nested logit model Dependent variable Choice Number of obs.= 210, skipped 0 obs --------+-------------------------------------------------------------------- |Model for Choice Among Alternatives AT| 2.38614*** .36950 6.46 .0000 1.66193 3.11035 AB| .76659** .32387 2.37 .0179 .13182 1.40136 BC| -.07659*** .01004 -7.63 .0000 -.09627 -.05691 |Model for Choice Among Branches AH| -.01386*** .00428 -3.24 .0012 -.02225 -.00548 AIV| .04165 .05691 .73 .4642 -.06989 .15319 --------+-------------------------------------------------------------------- And, this is what I got using the same command (FIML rather than LIML estimator): Hessian is not positive definite at start values. Error 803: Hessian is not positive definite at start values. B0 is too far from solution for Newton method. Iterative procedure has converged ----------------------------------------------------------------------------- Start values obtained using MNL model Dependent variable Choice Number of obs.= 210, skipped 0 obs --------+-------------------------------------------------------------------- | Standard Prob. 95% Confidence MODE| Coefficient Error z |z|>Z* Interval --------+-------------------------------------------------------------------- |Model for Choice Among Alternatives................................. AT| .67162*** .19719 3.41 .0007 .28513 1.05810 AB| -.32457 .....(Fixed Parameter)..... BC| -.01981*** .5015D-06 ******** .0000 -.01981 -.01981 |Model for Choice Among Branches..................................... AH| -.01539 .....(Fixed Parameter)..... AIV| 0.0 .....(Fixed Parameter)..... --------+-------------------------------------------------------------------- ----------------------------------------------------------------------------- FIML Nested Multinomial Logit Model Dependent variable MODE Warning: Model does not contain a full set of ASCs. R-sqrd is problematic. Number of obs.= 210, skipped 0 obs --------+-------------------------------------------------------------------- | Standard Prob. 95% Confidence MODE| Coefficient Error z |z|>Z* Interval --------+-------------------------------------------------------------------- |Attributes in the Utility Functions (beta).......................... AT| 2.36903*** .37054 6.39 .0000 1.64278 3.09528 AB| .76469** .32447 2.36 .0184 .12875 1.40064 BC| -.07685*** .01004 -7.65 .0000 -.09653 -.05717 |Attributes of Branch Choice Equations (alpha)....................... AH| .00473 .00681 .69 .4876 -.00863 .01808 AIV| 0.0 .9261D+14 .00 1.0000 -.18151D+15 .18151D+15 |IV parameters (RU1), tau(b|l,r),sigma(l|r),phi(r)................... FLY| .08202 .06278 1.31 .1914 -.04103 .20507 GROUND| -.04438 .06521 -.68 .4961 -.17220 .08343 --------+-------------------------------------------------------------------- I've tried with different specifications but, the Sequential estimation seems to be unavailable. Am I doing something wrong (I just copied and pasted the same commands)? Or Is the Sequential estimator unavailable in NLogit 6 (I am using the student license)? Best wishes, Saul Basurto From M.A.Young at soton.ac.uk Thu Jul 27 22:49:07 2017 From: M.A.Young at soton.ac.uk (Young M.A.) Date: Thu, 27 Jul 2017 12:49:07 +0000 Subject: [Limdep Nlogit List] Error components model - specify correlation based on variable Message-ID: <45211921C1E44145B3C8B129112470077662747A@SRV00360.soton.ac.uk> Hi, As far as I can tell from the NLOGIT manual, the only specification available for the error components model is the one that uses the tree structure (nested) analogy. I would like to specify the correlation based on a variable, which would seem to be what the error components interpretation of ML should allow. In my case I want this to be the distance between each pair of alternatives (or some function of this distance). Is this possible using NLOGIT? Thanks Marcus From richard.turner at imarketresearch.com Fri Jul 28 05:41:10 2017 From: richard.turner at imarketresearch.com (Richard Turner) Date: Thu, 27 Jul 2017 15:41:10 -0400 Subject: [Limdep Nlogit List] CLogit Likelihood Function 2nd Derivative - Clarification Message-ID: Greetings Community, I am trying to implement a *conditional logit* model in Excel to gain a better understanding of the estimation procedure. To my knowledge, the procedure for estimating *conditional logits* is the same as estimating *multinomial logits*, at least in terms of the probability equation, 1st derivatives for the Jacobian, and the *2nd partial derivatives for the Hessian*. After implementing the model in Excel, I still do not fully understand the *second partial derivatives that create the Hessian*. The second partial derivatives are: Where x_ik is the ith observation of the kth variable, ?_ik is the probability of the ith observation of the jth alternative. After reading pages 9-13 of this document , I am not sure when you would need to use the second derivative when j? is not equal to j. *Could someone explain, as simply as possible, when this derivative would be used?* In my Excel model, I have 3 alternatives and two variables, Var1 and Var2, which I represent as x1 and x2. I've implemented the newton-raphson algorithm to maximize the likelihood function using only the second derivative where j? equals to j. After several iterations, my Newton and Jacobian values were sufficiently equal to zero and my parameters equaled parameters estimated using NLOGIT. I assume, the model was estimated correctly using only the second derivative for the Hessian when j? equals to j, but I am still not sure when to use the other derivative when j? is not equal to j. For reference, I have attached my excel model Thank you very much for your time, ______________________________________________ *Richard Turner* *iMarketResearch* *Easier, Faster, Better Advanced Market Research* www.imarketresearch.com Address: 12 Penns Trail, Suite A, Newtown, PA 18940 From y2.ali at qut.edu.au Mon Jul 31 19:15:14 2017 From: y2.ali at qut.edu.au (Yasir Ali) Date: Mon, 31 Jul 2017 09:15:14 +0000 Subject: [Limdep Nlogit List] Help required on silly matter Message-ID: Hello, I am quite new to NLOGIT and trying to run MNL model. While running the model, I am getting this error "Error 1003: A choice label appears more than once in the tree spec." I am unable to figure out that where is the issue? Either in the code for running MNL or in the data? I have checked the data and in the time series data, only one choice is selected at one instance. Can anyone please help me to resolve this issue? Thank you. Cheers, Yasir From y2.ali at qut.edu.au Mon Jul 31 19:18:07 2017 From: y2.ali at qut.edu.au (Yasir Ali) Date: Mon, 31 Jul 2017 09:18:07 +0000 Subject: [Limdep Nlogit List] Help required on silly matter Message-ID: Hello, I am quite new to NLOGIT and trying to run MNL model. While running the model, I am getting this error "Error 1003: A choice label appears more than once in the tree spec." I am unable to figure out that where is the issue? Either in the code for running MNL or in the data? I have checked the data and in the time series data, only one choice is selected at one instance. Can anyone please help me to resolve this issue? Thank you. Cheers, Yasir