From doctorjasonong at gmail.com Wed Jan 1 12:27:48 2020 From: doctorjasonong at gmail.com (Jason Ong) Date: Wed, 1 Jan 2020 12:27:48 +1100 Subject: [Limdep Nlogit List] command for Variance-covariance matrix in NLOGIT Message-ID: Sorry if this question is too simple but I couldn't find it easily in the manual... I am running a latent class model and wanted to display the variance-covariance matrix too - what command should I use? (I want to get the covariance estimates between two coefficients) My code is below thanks for your help Best, Jason Ong |-> LCLOGIT ; Lhs = choicev, cset, altij ; Choices = A,B ; Pds = 16 ; pts=4;par ; lcm=infreq, student, manypart, migrant ; Model: U(A,B)=c20*cost1+c40*cost2+c60*cost3+sp20min*speed1+sp1day*speed2+sp3day*speed3 +win6w*window1+win12w*window2+Oral*mode1+FingerP*mode2+Ac95*accur1+Ac99*accur2+Ac999*accur3 +Cman*collect1+Cpeer*collect2$ *Jason Ong* Twitter: @DrJasonJOng PhD, MMed, MBBS, FAChSHM, FRACGP Sexual Health Physician, Melbourne Sexual Health Centre, Alfred Health Associate Professor (Hon), London School of Hygiene and Tropical Medicine, UK Central Clinical School, Monash University, Australia Melbourne School of Population and Global Health, University of Melbourne, Australia Associate Editor, Sexually Transmitted Infections Special Issues Editor, Sexual Health Board Director, ASHM (www.ashm.org.au) https://protect-au.mimecast.com/s/oWAKCnxyErCPvp0AU9WwF9?domain=lshtm.ac.uk https://protect-au.mimecast.com/s/gjn1CoVzGQigEpNjuzl3I9?domain=researchgate.net The race is not to the swift or the battle to the strong, nor does food come to the wise or wealth to the brilliant or favor to the learned; but time and chance happen to them all. From doctorjasonong at gmail.com Fri Jan 3 11:20:23 2020 From: doctorjasonong at gmail.com (Jason Ong) Date: Fri, 3 Jan 2020 11:20:23 +1100 Subject: [Limdep Nlogit List] Getting standard errors for omitted variables? Message-ID: hi, When I run an MNL model, it gives me outputs of the coefficients for each attribute level However, I would also like to get the standard errors (to calculate the 95% CI) for the omitted/reference level how could I calculate this? thank you kindly for your help Discrete choice (multinomial logit) model Dependent variable Choice Log likelihood function -3332.83194 Estimation based on N = 7008, K = 14 Inf.Cr.AIC = 6693.7 AIC/N = .955 --------------------------------------- Log likelihood R-sqrd R2Adj ASCs only model must be fit separately Use NLOGIT ;...;RHS=ONE$ Note: R-sqrd = 1 - logL/Logl(constants) Warning: Model does not contain a full set of ASCs. R-sqrd is problematic. Use model setup with ;RHS=one to get LogL0. --------------------------------------- Response data are given as ind. choices Number of obs.= 7008, skipped 0 obs --------+-------------------------------------------------------------------- | Standard Prob. 95% Confidence CHOICEV| Coefficient Error z |z|>Z* Interval --------+-------------------------------------------------------------------- C20|1| .40410*** .03059 13.21 .0000 .34413 .46406 C40|1| -.46829*** .03070 -15.26 .0000 -.52845 -.40813 C60|1| -1.26084*** .03594 -35.08 .0000 -1.33128 -1.19040 ACVEND|1| -.09467** .04822 -1.96 .0496 -.18918 -.00017 ACSHEL|1| .45141*** .05013 9.00 .0000 .35315 .54967 ACMED|1| .08191* .04740 1.73 .0840 -.01099 .17480 ACPHAR|1| .05906 .05280 1.12 .2633 -.04443 .16255 ACCBO|1| -.13720*** .05293 -2.59 .0095 -.24095 -.03345 ACSOPV|1| -.81479*** .05596 -14.56 .0000 -.92448 -.70510 PLARBR|1| -.10541*** .03289 -3.21 .0013 -.16986 -.04095 PSMPL|1| .04375 .03167 1.38 .1671 -.01832 .10582 PSMBR|1| .03257 .03254 1.00 .3168 -.03120 .09635 INFVID|1| -.06273** .02688 -2.33 .0196 -.11540 -.01005 INFCHA|1| -.05929** .02529 -2.34 .0191 -.10886 -.00972 --------+-------------------------------------------------------------------- *Jason Ong* Twitter: @DrJasonJOng PhD, MMed, MBBS, FAChSHM, FRACGP Sexual Health Physician, Melbourne Sexual Health Centre, Alfred Health Associate Professor (Hon), London School of Hygiene and Tropical Medicine, UK Central Clinical School, Monash University, Australia Melbourne School of Population and Global Health, University of Melbourne, Australia Associate Editor, Sexually Transmitted Infections Special Issues Editor, Sexual Health Board Director, ASHM (www.ashm.org.au) https://protect-au.mimecast.com/s/6NWaCZYM2VFk9JELizttba?domain=lshtm.ac.uk https://protect-au.mimecast.com/s/RrPzC1WZXriPNV1ohGBJsJ?domain=researchgate.net If you are more fortunate than others, build a longer table, not a taller fence. From zanoli at agrecon.univpm.it Fri Jan 3 11:31:11 2020 From: zanoli at agrecon.univpm.it (Raffaele Zanoli) Date: Fri, 3 Jan 2020 01:31:11 +0100 Subject: [Limdep Nlogit List] Bounded coefficients in NLOGIT estimation Message-ID: Hello Bill, I am trying to estimate a model in NLOGIT6 (for noe a simple MNL) where I have an interaction coefficient that should be bounded between the unit interval [0,1]. Is there any way to do this? I could not find any mention to this in the Manuals (only fixed chefs or starting values) Thanks Raffaele Prof. Dr. Raffaele Zanoli Professor of Food Marketing & Management Department of Agricultural, Food and Environmental Sciences Universit? Politecnica delle Marche Via Brecce Bianche 60129 Ancona- Italy tel. +39 071 220 4929 mobile +39 3483235654 From wgreene at stern.nyu.edu Fri Jan 3 12:13:26 2020 From: wgreene at stern.nyu.edu (William Greene) Date: Thu, 2 Jan 2020 20:13:26 -0500 Subject: [Limdep Nlogit List] Getting standard errors for omitted variables? In-Reply-To: References: Message-ID: Jason. The coefficient on the base level is zero. There is no standard error for it. The zero is a fixed value, not an estimated coefficient. /Bill Greene On Thu, Jan 2, 2020 at 7:21 PM Jason Ong via Limdep < limdep at mailman.sydney.edu.au> wrote: > hi, > > When I run an MNL model, it gives me outputs of the coefficients for each > attribute level > However, I would also like to get the standard errors (to calculate the 95% > CI) for the omitted/reference level > how could I calculate this? > thank you kindly for your help > > Discrete choice (multinomial logit) model > > Dependent variable Choice > > Log likelihood function -3332.83194 > > Estimation based on N = 7008, K = 14 > > Inf.Cr.AIC = 6693.7 AIC/N = .955 > > --------------------------------------- > > Log likelihood R-sqrd R2Adj > > ASCs only model must be fit separately > > Use NLOGIT ;...;RHS=ONE$ > > Note: R-sqrd = 1 - logL/Logl(constants) > > Warning: Model does not contain a full > > set of ASCs. R-sqrd is problematic. Use > > model setup with ;RHS=one to get LogL0. > > --------------------------------------- > > Response data are given as ind. choices > > Number of obs.= 7008, skipped 0 obs > > > --------+-------------------------------------------------------------------- > > | Standard Prob. 95% Confidence > > CHOICEV| Coefficient Error z |z|>Z* Interval > > > --------+-------------------------------------------------------------------- > > C20|1| .40410*** .03059 13.21 .0000 .34413 .46406 > > C40|1| -.46829*** .03070 -15.26 .0000 -.52845 -.40813 > > C60|1| -1.26084*** .03594 -35.08 .0000 -1.33128 -1.19040 > > ACVEND|1| -.09467** .04822 -1.96 .0496 -.18918 -.00017 > > ACSHEL|1| .45141*** .05013 9.00 .0000 .35315 .54967 > > ACMED|1| .08191* .04740 1.73 .0840 -.01099 .17480 > > ACPHAR|1| .05906 .05280 1.12 .2633 -.04443 .16255 > > ACCBO|1| -.13720*** .05293 -2.59 .0095 -.24095 -.03345 > > ACSOPV|1| -.81479*** .05596 -14.56 .0000 -.92448 -.70510 > > PLARBR|1| -.10541*** .03289 -3.21 .0013 -.16986 -.04095 > > PSMPL|1| .04375 .03167 1.38 .1671 -.01832 .10582 > > PSMBR|1| .03257 .03254 1.00 .3168 -.03120 .09635 > > INFVID|1| -.06273** .02688 -2.33 .0196 -.11540 -.01005 > > INFCHA|1| -.05929** .02529 -2.34 .0191 -.10886 -.00972 > > > --------+-------------------------------------------------------------------- > > > *Jason Ong* > Twitter: @DrJasonJOng > PhD, MMed, MBBS, FAChSHM, FRACGP > > Sexual Health Physician, Melbourne Sexual Health Centre, Alfred Health > Associate Professor (Hon), London School of Hygiene and Tropical Medicine, > UK > Central Clinical School, Monash University, Australia > Melbourne School of Population and Global Health, University of Melbourne, > Australia > Associate Editor, Sexually Transmitted Infections > Special Issues Editor, Sexual Health > Board Director, ASHM (https://protect-au.mimecast.com/s/fsv6CNLwM9iO3rV7im9LZS?domain=ashm.org.au) > https://protect-au.mimecast.com/s/diKNCOMxNyt7x95nUvIDkj?domain=lshtm.ac.uk > https://protect-au.mimecast.com/s/THlzCP7yOZtjXrvWS118F3?domain=researchgate.net > > If you are more fortunate than others, build a longer table, not a taller > fence. > _______________________________________________ > Limdep site list > Limdep at mailman.sydney.edu.au > http://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/qZTdCQnzP0t2PEBOtklrlP?domain=people.stern.nyu.edu Email: wgreene at stern.nyu.edu Ph. +1.646.596.3296 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 From doctorjasonong at gmail.com Fri Jan 3 12:20:46 2020 From: doctorjasonong at gmail.com (Jason Ong) Date: Fri, 3 Jan 2020 12:20:46 +1100 Subject: [Limdep Nlogit List] Getting standard errors for omitted variables? In-Reply-To: References: Message-ID: thanks Bill for the prompt response Sorry I should have clarified that this is effects coded. so my coefficient for my reference level (say for the first attribute) will be -1* (0.4 + (-0.5) + (-1.3)) = 1.4 is there any way to calculate the standard error around this 1.4 value? *Jason Ong* Twitter: @DrJasonJOng PhD, MMed, MBBS, FAChSHM, FRACGP Sexual Health Physician, Melbourne Sexual Health Centre, Alfred Health Associate Professor (Hon), London School of Hygiene and Tropical Medicine, UK Central Clinical School, Monash University, Australia Melbourne School of Population and Global Health, University of Melbourne, Australia Associate Editor, Sexually Transmitted Infections Special Issues Editor, Sexual Health Board Director, ASHM (https://protect-au.mimecast.com/s/ejWVCVAGXPtWyL8guGGc2v?domain=ashm.org.au) https://protect-au.mimecast.com/s/zfnYCWLJY7iMqg8DuxIWvs?domain=lshtm.ac.uk https://protect-au.mimecast.com/s/s6KnCXLKZoi8KyZOIDVyRD?domain=researchgate.net If you are more fortunate than others, build a longer table, not a taller fence. On Fri, Jan 3, 2020 at 12:13 PM William Greene via Limdep < limdep at mailman.sydney.edu.au> wrote: > Jason. The coefficient on the base level is zero. There is no standard > error for > it. The zero is a fixed value, not an estimated coefficient. > /Bill Greene > > On Thu, Jan 2, 2020 at 7:21 PM Jason Ong via Limdep < > limdep at mailman.sydney.edu.au> wrote: > > > hi, > > > > When I run an MNL model, it gives me outputs of the coefficients for each > > attribute level > > However, I would also like to get the standard errors (to calculate the > 95% > > CI) for the omitted/reference level > > how could I calculate this? > > thank you kindly for your help > > > > Discrete choice (multinomial logit) model > > > > Dependent variable Choice > > > > Log likelihood function -3332.83194 > > > > Estimation based on N = 7008, K = 14 > > > > Inf.Cr.AIC = 6693.7 AIC/N = .955 > > > > --------------------------------------- > > > > Log likelihood R-sqrd R2Adj > > > > ASCs only model must be fit separately > > > > Use NLOGIT ;...;RHS=ONE$ > > > > Note: R-sqrd = 1 - logL/Logl(constants) > > > > Warning: Model does not contain a full > > > > set of ASCs. R-sqrd is problematic. Use > > > > model setup with ;RHS=one to get LogL0. > > > > --------------------------------------- > > > > Response data are given as ind. choices > > > > Number of obs.= 7008, skipped 0 obs > > > > > > > --------+-------------------------------------------------------------------- > > > > | Standard Prob. 95% Confidence > > > > CHOICEV| Coefficient Error z |z|>Z* Interval > > > > > > > --------+-------------------------------------------------------------------- > > > > C20|1| .40410*** .03059 13.21 .0000 .34413 .46406 > > > > C40|1| -.46829*** .03070 -15.26 .0000 -.52845 -.40813 > > > > C60|1| -1.26084*** .03594 -35.08 .0000 -1.33128 -1.19040 > > > > ACVEND|1| -.09467** .04822 -1.96 .0496 -.18918 -.00017 > > > > ACSHEL|1| .45141*** .05013 9.00 .0000 .35315 .54967 > > > > ACMED|1| .08191* .04740 1.73 .0840 -.01099 .17480 > > > > ACPHAR|1| .05906 .05280 1.12 .2633 -.04443 .16255 > > > > ACCBO|1| -.13720*** .05293 -2.59 .0095 -.24095 -.03345 > > > > ACSOPV|1| -.81479*** .05596 -14.56 .0000 -.92448 -.70510 > > > > PLARBR|1| -.10541*** .03289 -3.21 .0013 -.16986 -.04095 > > > > PSMPL|1| .04375 .03167 1.38 .1671 -.01832 .10582 > > > > PSMBR|1| .03257 .03254 1.00 .3168 -.03120 .09635 > > > > INFVID|1| -.06273** .02688 -2.33 .0196 -.11540 -.01005 > > > > INFCHA|1| -.05929** .02529 -2.34 .0191 -.10886 -.00972 > > > > > > > --------+-------------------------------------------------------------------- > > > > > > *Jason Ong* > > Twitter: @DrJasonJOng > > PhD, MMed, MBBS, FAChSHM, FRACGP > > > > Sexual Health Physician, Melbourne Sexual Health Centre, Alfred Health > > Associate Professor (Hon), London School of Hygiene and Tropical > Medicine, > > UK > > Central Clinical School, Monash University, Australia > > Melbourne School of Population and Global Health, University of > Melbourne, > > Australia > > Associate Editor, Sexually Transmitted Infections > > Special Issues Editor, Sexual Health > > Board Director, ASHM (https://protect-au.mimecast.com/s/Q1DBCYWL1vinlJXASVPz8X?domain=ashm.org.au) > > https://protect-au.mimecast.com/s/zfnYCWLJY7iMqg8DuxIWvs?domain=lshtm.ac.uk > > https://protect-au.mimecast.com/s/s6KnCXLKZoi8KyZOIDVyRD?domain=researchgate.net > > > > If you are more fortunate than others, build a longer table, not a taller > > fence. > > _______________________________________________ > > Limdep site list > > Limdep at mailman.sydney.edu.au > > http://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/IkZyCZYM2VFkgVYQsxTtr3?domain=people.stern.nyu.edu > Email: wgreene at stern.nyu.edu > Ph. +1.646.596.3296 > 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 > _______________________________________________ > Limdep site list > Limdep at mailman.sydney.edu.au > http://limdep.itls.usyd.edu.au > > From wgreene at stern.nyu.edu Fri Jan 3 13:09:45 2020 From: wgreene at stern.nyu.edu (William Greene) Date: Thu, 2 Jan 2020 21:09:45 -0500 Subject: [Limdep Nlogit List] Bounded coefficients in NLOGIT estimation In-Reply-To: References: Message-ID: Raffaele. I don't see clearly what "an interaction coefficient" looks like in a multinomial choice model where there are multiple utility functions. In any event, there is no direct way to constrain the coefficients in the MNL. I would note, in general, when you try to constrain the coefficients to a range, one of two things happens. (1) The constraint is unnecessary, and the unrestricted coefficient ends up in the right way anyway. (2) The constraint is inappropriate for the data, end the estimates end up at a boundary of the range. What does not happen is that there is some constrained estimate in the middle of the range that would not be found if some constraint were not imposed. In my experience, there is usually no payoff to trying to constrain a parameter to a range like this. Regards, Bill On Thu, Jan 2, 2020 at 7:31 PM Raffaele Zanoli wrote: > Hello Bill, > > I am trying to estimate a model in NLOGIT6 (for noe a simple MNL) where I > have an interaction coefficient that should be bounded between the unit > interval [0,1]. > > Is there any way to do this? I could not find any mention to this in the > Manuals (only fixed chefs or starting values) > > Thanks > > Raffaele > > > Prof. Dr. Raffaele Zanoli > Professor of Food Marketing & Management > Department of Agricultural, Food and Environmental Sciences > Universit? Politecnica delle Marche > Via Brecce Bianche > 60129 Ancona- Italy > tel. +39 071 220 4929 > mobile +39 3483235654 > > _______________________________________________ > Limdep site list > Limdep at mailman.sydney.edu.au > http://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/eJWdCL7rK8tG9kXPuBzmb3?domain=people.stern.nyu.edu Email: wgreene at stern.nyu.edu Ph. +1.646.596.3296 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 From doctorjasonong at gmail.com Wed Jan 8 18:28:50 2020 From: doctorjasonong at gmail.com (Jason Ong) Date: Wed, 8 Jan 2020 18:28:50 +1100 Subject: [Limdep Nlogit List] majority of data now skipped in MNL model In-Reply-To: References: Message-ID: kindly ignore the email I just sent It's rather strange... I just restarted my NLOGIT and tried to run the models again and it was back to normal (i.e. used all my data) sorry to trouble you. *Jason Ong* Twitter: @DrJasonJOng PhD, MMed, MBBS, FAChSHM, FRACGP Sexual Health Physician, Melbourne Sexual Health Centre, Alfred Health Associate Professor (Hon), London School of Hygiene and Tropical Medicine, UK Central Clinical School, Monash University, Australia Melbourne School of Population and Global Health, University of Melbourne, Australia Associate Editor, Sexually Transmitted Infections Special Issues Editor, Sexual Health Board Director, ASHM (www.ashm.org.au) https://protect-au.mimecast.com/s/2SbtCmOxDQtgwnLxIGQghp?domain=lshtm.ac.uk https://protect-au.mimecast.com/s/xRzPCnxyErCP0opEiJjYT6?domain=researchgate.net If you are more fortunate than others, build a longer table, not a taller fence. On Wed, Jan 8, 2020 at 6:19 PM Jason Ong wrote: > hi, > > I am not sure what happened to my dataset > A few months ago when I ran it, it used all my data points > but when I tried to run an MNL model again it came up with this error > > +------------------------------------------------------+ > |WARNING: Bad observations were found in the sample. | > |Found 2616 bad observations among 2808 individuals. | > |You can use ;CheckData to get a list of these points. | > +------------------------------------------------------+ > > when I used ;CheckData it came up with a long list (pasting the last few) > - but I am not sure what the es.') means? > > 8368 2790 es.') > 8371 2791 es.') > 8374 2792 es.') > 8377 2793 es.') > 8380 2794 es.') > 8383 2795 es.') > 8386 2796 es.') > 8389 2797 es.') > 8392 2798 es.') > 8395 2799 es.') > 8398 2800 es.') > 8401 2801 es.') > 8404 2802 es.') > 8407 2803 es.') > 8410 2804 es.') > 8413 2805 es.') > 8416 2806 es.') > 8419 2807 es.') > 8422 2808 es.') > > Below is my MNL code and my database attached > hope you can shed some light on this mystery... > thanks. > > NLOGIT > ; Lhs = choicev, cset, altij > ; Choices = A,B,C > ?;CheckData > ; Model: > U(A,B)=cost1*cost1+cost2*cost2+cost3*cost3+who*who+swab*swab > > +extra1*extra1+extra2*extra2+extra3*extra3+supp1*supp1+supp2*supp2+supp3*supp3 > +loc1*loc1+loc2*loc2+loc3*loc3+loc4*loc4+loc5*loc5+loc6*loc6+loc7*loc7/ > U(C)=neither$ > > > *Jason Ong* > Twitter: @DrJasonJOng > PhD, MMed, MBBS, FAChSHM, FRACGP > > Sexual Health Physician, Melbourne Sexual Health Centre, Alfred Health > Associate Professor (Hon), London School of Hygiene and Tropical Medicine, > UK > Central Clinical School, Monash University, Australia > Melbourne School of Population and Global Health, University of Melbourne, > Australia > Associate Editor, Sexually Transmitted Infections > Special Issues Editor, Sexual Health > Board Director, ASHM (www.ashm.org.au) > https://protect-au.mimecast.com/s/2SbtCmOxDQtgwnLxIGQghp?domain=lshtm.ac.uk > https://protect-au.mimecast.com/s/xRzPCnxyErCP0opEiJjYT6?domain=researchgate.net > > If you are more fortunate than others, build a longer table, not a taller > fence. > > From bills at gra-inc.com Wed Jan 15 08:00:40 2020 From: bills at gra-inc.com (William Spitz) Date: Tue, 14 Jan 2020 16:00:40 -0500 Subject: [Limdep Nlogit List] Hausman-Taylor estimator for panel data Message-ID: <06000f11-365d-d311-6efa-81f2c1038a97@gra-inc.com> Hi. I'm trying to the Hausman-Taylor estimator. There are four sets of variables in the model: x1 = time-varying uncorrelated with u[i] x2 = time-varying correlated with u[i] f1 = time-invariant uncorrelated with u[i] f2 = time-invariant correlated with u[i] The manual claims that only the first of these required. For my purposes I have only x1 and x2 defined; I have no time-invariant variables, but I still want to use H-T estimator to account for the x2 variables that (I suspect) are highly correlated with the group effects. But when I issue the final command to get the H-T estimates, I get an error message stating: No time invariant vars uncorrelated with u(i) No time invariant vars correlated with u(i) How can I get H-T estimates without time-invariant variables? -- William Spitz From wgreene at stern.nyu.edu Fri Jan 17 06:24:42 2020 From: wgreene at stern.nyu.edu (William Greene) Date: Thu, 16 Jan 2020 14:24:42 -0500 Subject: [Limdep Nlogit List] Hausman-Taylor estimator for panel data In-Reply-To: <06000f11-365d-d311-6efa-81f2c1038a97@gra-inc.com> References: <06000f11-365d-d311-6efa-81f2c1038a97@gra-inc.com> Message-ID: Bill. If you have no time invariant observed data, then you can just use 2SLS or a random effects version, 2SRE. H&T is written for models that contain time invariant effects (observed and unobserved). Regards Bill Greene On Tue, Jan 14, 2020 at 4:01 PM William Spitz wrote: > Hi. I'm trying to the Hausman-Taylor estimator. There are four sets of > variables in the model: > x1 = time-varying uncorrelated with u[i] > x2 = time-varying correlated with u[i] > f1 = time-invariant uncorrelated with u[i] > f2 = time-invariant correlated with u[i] > > The manual claims that only the first of these required. For my purposes > I have only x1 and x2 defined; I have no time-invariant variables, but I > still want to use H-T estimator to account for the x2 variables that (I > suspect) are highly correlated with the group effects. > But when I issue the final command to get the H-T estimates, I get an > error message stating: > No time invariant vars uncorrelated with u(i) > No time invariant vars correlated with u(i) > > How can I get H-T estimates without time-invariant variables? > > -- > William Spitz > > _______________________________________________ > Limdep site list > Limdep at mailman.sydney.edu.au > http://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/M_x5C1WZXriQjVA9TLYNG1?domain=people.stern.nyu.edu Email: wgreene at stern.nyu.edu Ph. +1.646.596.3296 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 From kjreynoldsatp at icloud.com Tue Jan 21 03:32:55 2020 From: kjreynoldsatp at icloud.com (KENNETH REYNOLDS) Date: Mon, 20 Jan 2020 16:32:55 -0000 Subject: [Limdep Nlogit List] Allocating memory does not work Message-ID: Has anyone else had this issue, and if so, what was the fix?? thank you. -- I am having the following problem allocating enough memory to LIMDEP/NLOGIT6 in order to run regressions. The data set that I am trying to load contains 24 variables (columns) and 275,875 observations (rows). I have used import variables and read commands, but the results are the same, not enough memory to load the data set. I am not a new NLOGIT6/LIMDEP user, and I have run large data sets with older versions; however, I was able to set both the cell numbers in memory allocation and the rows in dimensions, and then the number of columns would adjust automatically. This is not happening in my version of NLOGIT6 (Sep 7, 2016). I can only set the number of required cells in memory allocation, but I cannot access either of the dimensions - rows or columns. Thus I can set the number of cells to my requirement of 6,621,000; however, the number of columns remain at 900 while the number of rows adjust to 7,356 - I cannot manually change the number of rows. The rest of the rows in my data set are then set to missing data. I then receive errors 522?(READ needs 6621000 cells. NKMAX= 5000000. Use Project menu ...) and 7 (READ - Data set is too large. Expand data area (Project...). Attempting to reset the memory allocation through project/settings or tools/options/projects is not helpful. Any help would be most appreciated. Thank you, Ken ? From wgreene at stern.nyu.edu Tue Jan 21 05:22:20 2020 From: wgreene at stern.nyu.edu (William Greene) Date: Mon, 20 Jan 2020 13:22:20 -0500 Subject: [Limdep Nlogit List] Allocating memory does not work In-Reply-To: References: Message-ID: Ken. tools/options/projects is the way to reset this value. 1. Start the program. 2. Go directly to tools/options/projects 3. Set the number of cells to 10 million. Don't set the number of rows, just the number of cells. 4. Press OK and leaves the tools menu. 5. Exit the program without reading the data. 6. Restart the program. 7. READ or IMPORT the data. 8. Use File:Save Project As to save the project. Now continue your processing. Before you exit the program, save the project. Then, use File->Open Project to resume the session. .Bill Greene On Mon, Jan 20, 2020 at 11:33 AM KENNETH REYNOLDS via Limdep < limdep at mailman.sydney.edu.au> wrote: > Has anyone else had this issue, and if so, what was the fix? thank you. > -- > > I am having the following problem allocating enough memory to > LIMDEP/NLOGIT6 in order to run regressions. The data set that I am trying > to load contains 24 variables (columns) and 275,875 observations (rows). I > have used import variables and read commands, but the results are the same, > not enough memory to load the data set. I am not a new NLOGIT6/LIMDEP user, > and I have run large data sets with older versions; however, I was able to > set both the cell numbers in memory allocation and the rows in dimensions, > and then the number of columns would adjust automatically. > > > > > This is not happening in my version of NLOGIT6 (Sep 7, 2016). I can only > set the number of required cells in memory allocation, but I cannot access > either of the dimensions - rows or columns. Thus I can set the number of > cells to my requirement of 6,621,000; however, the number of columns remain > at 900 while the number of rows adjust to 7,356 - I cannot manually change > the number of rows. The rest of the rows in my data set are then set to > missing data. > > > > > I then receive errors 522 (READ needs 6621000 cells. NKMAX= 5000000. Use > Project menu ...) and 7 (READ - Data set is too large. Expand data area > (Project...). > > > > > Attempting to reset the memory allocation through project/settings or > tools/options/projects is not helpful. > > > > > Any help would be most appreciated. > > > > > Thank you, Ken > > > _______________________________________________ > Limdep site list > Limdep at mailman.sydney.edu.au > http://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/PeIbCWLJY7in46Kvt6mQck?domain=people.stern.nyu.edu Email: wgreene at stern.nyu.edu Ph. +1.646.596.3296 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 From kjreynoldsatp at icloud.com Tue Jan 21 08:57:02 2020 From: kjreynoldsatp at icloud.com (KENNETH REYNOLDS) Date: Mon, 20 Jan 2020 21:57:02 -0000 Subject: [Limdep Nlogit List] Allocating memory does not work Message-ID: Thanks Bill, it worked! I did this before, but unfortunately I left out one step.? Sigh ... Happy New Year Bill! On January 20, 2020 at 1:23 PM, William Greene via Limdep wrote: Ken. tools/options/projects is the way to reset this value. 1. Start the program. 2. Go directly to tools/options/projects 3. Set the number of cells to 10 million. Don't set the number of rows, just the number of cells. 4. Press OK and leaves the tools menu. 5. Exit the program without reading the data. 6. Restart the program. 7. READ or IMPORT the data. 8. Use File:Save Project As to save the project. Now continue your processing. Before you exit the program, save the project. Then, use File->Open Project to resume the session. .Bill Greene On Mon, Jan 20, 2020 at 11:33 AM KENNETH REYNOLDS via Limdep < limdep at mailman.sydney.edu.au> wrote: Has anyone else had this issue, and if so, what was the fix? thank you. -- I am having the following problem allocating enough memory to LIMDEP/NLOGIT6 in order to run regressions. The data set that I am trying to load contains 24 variables (columns) and 275,875 observations (rows). I have used import variables and read commands, but the results are the same, not enough memory to load the data set. I am not a new NLOGIT6/LIMDEP user, and I have run large data sets with older versions; however, I was able to set both the cell numbers in memory allocation and the rows in dimensions, and then the number of columns would adjust automatically. This is not happening in my version of NLOGIT6 (Sep 7, 2016). I can only set the number of required cells in memory allocation, but I cannot access either of the dimensions - rows or columns. Thus I can set the number of cells to my requirement of 6,621,000; however, the number of columns remain at 900 while the number of rows adjust to 7,356 - I cannot manually change the number of rows. The rest of the rows in my data set are then set to missing data. I then receive errors 522 (READ needs 6621000 cells. NKMAX= 5000000. Use Project menu ...) and 7 (READ - Data set is too large. Expand data area (Project...). Attempting to reset the memory allocation through project/settings or tools/options/projects is not helpful. Any help would be most appreciated. Thank you, Ken _______________________________________________ Limdep site list Limdep at mailman.sydney.edu.au http://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/JawsC2xZYvCDmgBrhn2_KQ?domain=people.stern.nyu.edu Email: wgreene at stern.nyu.edu Ph. +1.646.596.3296 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 _______________________________________________ Limdep site list Limdep at mailman.sydney.edu.au http://limdep.itls.usyd.edu.au