From Charles.Raux at laet.ish-lyon.cnrs.fr Fri Nov 3 02:19:36 2017 From: Charles.Raux at laet.ish-lyon.cnrs.fr (RAUX Charles) Date: Thu, 2 Nov 2017 15:19:36 +0000 Subject: [Limdep Nlogit List] nlogit, crosstab, list Message-ID: <72D6B3FEC474FA40B4981EDD2D672981CD40F6E8@CNMB02WVP.core-res.rootcore.local> Hello, I am trying various options of nlogit and getting weird or no results. I try nlogit ;choices=air, car, bus, train, none ;lhs=choicem ;model:U(air, car, bus, train, none)= +*duration+pc*price ;show model ;describe ;crosstab ;keep=pp$ The model estimates fine, the option keep too. But with crosstab I get only zeros in XTAB_PRB and XTAB_FRQ. When I add "list" the predictions are listed PREDICTED PROBABILITIES (* marks actual, + marks prediction.) Indiv AIR CAR BUS TRAIN NONE 1 .4223*+ .3993 .0686 .0593 .0505 2 .2120 .0752 .0054 .6232*+ .0842 ... But I get an error message "Error 803: Hessian is not positive definite at start values." And some parameters are fixed. This error doesn't happen when I don't use the option "list". What is wrong with these two issues? Charles Raux From Lixian.Qian at xjtlu.edu.cn Mon Nov 6 00:56:21 2017 From: Lixian.Qian at xjtlu.edu.cn (Lixian Qian) Date: Sun, 5 Nov 2017 13:56:21 +0000 Subject: [Limdep Nlogit List] Line search does not improve the function. Message-ID: Dear professors, I am running a Nested Logit Model in NLogit v5.0. I find the location of alternative specific constants (ASCs) in the utility function can impact the model convergence. When I specify the utility function with 3 alternatives as: U(A) = a1*X1 + a2*X2 / U(B) = a1*X1 + a2*X2 + asc_b + age_b*Age + sex_b *Gender / U(C) = a1*X1 + a2*X2 + asc_c + age_c*Age + sex_c *Gender The model estimation shows that Line search at iteration 61 does not improve fn. Exiting optimization. However, when I make a minor change on the utility functions to U(A) = a1*X1 + a2*X2 / U(B) = asc_b+ a1*X1 + a2*X2 + age_b*Age + sex_b *Gender / U(C) = asc_c + a1*X1 + a2*X2 + age_c*Age + sex_c *Gender The model can successfully converge, with the exact same LL function? all estimated coefficients as well as standard errors as in the final results (though not converged as mentioned in output). I don?t know what is wrong with the first case. Thank you and appreciate your advices. Best, Lixian From Charles.Raux at laet.ish-lyon.cnrs.fr Sat Nov 25 04:49:58 2017 From: Charles.Raux at laet.ish-lyon.cnrs.fr (RAUX Charles) Date: Fri, 24 Nov 2017 17:49:58 +0000 Subject: [Limdep Nlogit List] wtp in lclogit Message-ID: <72D6B3FEC474FA40B4981EDD2D672981CD423125@CNMB02WVP.core-res.rootcore.local> Hello, I have a specification of latent class multinomial model (below) lclogit; choices=air, car, bus, train, none ; lhs=choicem ; model: U(air, car, bus, train) = + * duration + pc * price / U(none) = anone ; show model ; pds=6; pts=2 ; wtp=da/pc, dc/pc, db/pc, dt/pc$ This works fine, parameters are significant and the matrix wtpi is created with four columns and nearly 300 rows (the individuals). But filled only with zero. What is wrong? Charles Raux From liberini at kof.ethz.ch Mon Nov 27 22:21:31 2017 From: liberini at kof.ethz.ch (Liberini Federica) Date: Mon, 27 Nov 2017 11:21:31 +0000 Subject: [Limdep Nlogit List] Standard Errors for Elasticities of Random Parameter Conditional Logit Message-ID: <213F594826F3B9428143CEB7C51517E04C73A7E4@MBX13.d.ethz.ch> Hello, I have estimated a random parameter conditional logit model, using the command NLOGIT. So far, I used the option ";Effects: x (*)" to obtain the elasticities of choosing each model's alternative w.r.t to variable x. Is there a way of obtaining standard errors for these elasticities? I tried using the command "Partial Effects", but Limpdep returns the error message saying that "Previous model did not store model setup for PARTIALS.". Many Thanks, Federica p.s.> here's a synthesis of the command I am using now nlogit; lhs=choice,nij,altcode ; choices=AT,BE,BR,CA,CH,DE,DK,DOM,ES,FI,FR,GB,IE,IT,NL,NO,RU,SE,US ; model: U(AT,BE,BR,CA,CH,DE,DK,DOM,ES,FI,FR,GB,IE,IT,NL,NO,RU,SE,US)= + * X1 + * X2 + *Dumm1*X2 + <0,0,0,0,0,0,0,domtx,0,0,0,0,0,0,0,0,0,0,0>*X2 + <0,0,0,0,0,0,0,mndomtx,0,0,0,0,0,0,0,0,0,0,0>*Dumm1*X2 + ; fcn= cbtx(n) ; rpl ; maxit=400 ; halton; pts=50 ; alg=bfgs ; effects: X2 (*) $