From amz at albertomzanni.com Wed Jul 1 00:04:27 2020
From: amz at albertomzanni.com (Alberto M Zanni)
Date: Tue, 30 Jun 2020 15:04:27 +0100
Subject: [Limdep Nlogit List] problem exporting results into Excel
Message-ID: <7222f047-249d-416e-9f5b-fa6211e57b1d@www.fastmail.com>
Dear all,
I have been struggling a bit while exporting results to excel (I am using NLogit version 5). This is a simple case
my command
open; export="C:\Users\abzn\Documents\Alberto\DLP\resultsMNL11.csv" $
nlogit;
lhs=choice;
choices=A1,A2,A3;
rhs=gcov, flower, tcov, isl_mid, isl_wide, bl_div,bl_slow, p_island, lane, tax, sq;
export $
The results from the output below are different from those in the Excel table produced by the command above as the confidence interval values are inserted as cofficients of the next variable and so on...
I found a previous message reporting this but could not find the reply as well. Am I doing anything wrong? is there a way to resolve this? or do you know of a quicker trick to copy results into excel?
thank you very much
Alberto
--------+--------------------------------------------------------------------
| Standard Prob. 95% Confidence
CHOICE| Coefficient Error z |z|>Z* Interval
--------+--------------------------------------------------------------------
GCOV| .01168*** .00266 4.40 .0000 .00647 .01688
FLOWER| .10408 .14518 .72 .4734 -.18047 .38862
TCOV| .00444 .00274 1.62 .1054 -.00093 .00981
ISL_MID| -.28088* .16282 -1.73 .0845 -.59999 .03824
ISL_WIDE| -.17648 .15446 -1.14 .2532 -.47921 .12625
BL_DIV| .25986** .12975 2.00 .0452 .00555 .51417
BL_SLOW| .01951 .12755 .15 .8785 -.23049 .26950
P_ISLAND| .11702 .09473 1.24 .2167 -.06866 .30270
LANE| -.12308** .05929 -2.08 .0379 -.23929 -.00687
TAX| -.00014*** .2962D-04 -4.69 .0000 -.00020 -.00008
SQ| -1.69585*** .31846 -5.33 .0000 -2.32002 -1.07169
--------+--------------------------------------------------------------------
Note: nnnnn.D-xx or D+xx => multiply by 10 to -xx or +xx.
Note: ***, **, * ==> Significance at 1%, 5%, 10% level.
-----------------------------------------------------------------------------
Last Model Estimation Results
Variable
Coeff.
Std.Err.
t-ratio
P-value
GCOV
1.17E-02
2.66E-03
4.39608
1.10E-05
FLOWER
6.47E-03
1.69E-02
0.104077
0.145178
TCOV
0.716888
0.473443
-0.18047
0.388621
ISL_MID
4.44E-03
2.74E-03
1.61915
0.105416
ISL_WIDE
-9.34E-04
9.81E-03
-0.28088
0.162818
BL_DIV
-1.72509
8.45E-02
-0.59999
3.82E-02
BL_SLOW
-0.17648
0.154456
-1.14258
0.253214
P_ISLAND
-0.47921
0.12625
0.25986
0.129754
LANE
2.00271
4.52E-02
5.55E-03
0.514174
TAX
1.95E-02
0.127552
0.152926
0.878457
SQ
-0.23049
0.269504
0.117019
9.47E-02
From Thao.T.Thai at monash.edu Thu Jul 2 13:32:24 2020
From: Thao.T.Thai at monash.edu (Thao Thai)
Date: Thu, 2 Jul 2020 13:32:24 +1000
Subject: [Limdep Nlogit List] MX logit using different algorithms
Message-ID:
Hi Nlogit users,
I am running a Mixed logit model with different algorithms which produce
exactly the same set of results.
- Newton-Raphson algorithm: I got this message "Line search at iteration 80
does not improve the function. Exiting optimization"
- BHHH algorithm: The model can converge after 80 iterations.
- BFGS algorithm: The model can converge after 81 iterations
Can I use the results given that three algorithms produced the same
coefficient estimates? Why do all three algorithms produce the same
results with different messages?
Thank you so much for your help so far. I really appreciate it.
Best regards,
Thao
From wgreene at stern.nyu.edu Fri Jul 3 06:27:13 2020
From: wgreene at stern.nyu.edu (William Greene)
Date: Thu, 2 Jul 2020 16:27:13 -0400
Subject: [Limdep Nlogit List] MX logit using different algorithms
In-Reply-To:
References:
Message-ID:
Thao. The mixed logit defaults to BFGS, which is the best algorithm for
that class of models.
The diagnostic in each case corresponds to the request that you issued, but
you should be
receiving only the BFGS results. You can check this by adding ;Output=3 to
the commands.
/Bill Greene
On Wed, Jul 1, 2020 at 11:33 PM Thao Thai via Limdep <
limdep at mailman.sydney.edu.au> wrote:
> Hi Nlogit users,
>
> I am running a Mixed logit model with different algorithms which produce
> exactly the same set of results.
>
> - Newton-Raphson algorithm: I got this message "Line search at iteration 80
> does not improve the function. Exiting optimization"
>
> - BHHH algorithm: The model can converge after 80 iterations.
>
> - BFGS algorithm: The model can converge after 81 iterations
>
> Can I use the results given that three algorithms produced the same
> coefficient estimates? Why do all three algorithms produce the same
> results with different messages?
>
> Thank you so much for your help so far. I really appreciate it.
> Best regards,
> Thao
> _______________________________________________
> 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/_wKdC81V0PTLKw64tnt5R4?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 Thao.T.Thai at monash.edu Sat Jul 4 12:29:42 2020
From: Thao.T.Thai at monash.edu (Thao Thai)
Date: Sat, 4 Jul 2020 12:29:42 +1000
Subject: [Limdep Nlogit List] Simulation using estimates from SP data and
ASCs from RP data
Message-ID:
Hi Nlogit users,
I am simulating some scenarios using the coefficient estimates from a CL
model on SP data and alternative specific constants (ASCs) calibrated to
reflect the RP data.
However, the reported market shares in the base case shows the market
shares observed in SP data. The reported market shares of the simulated
scenarios are also compared to these SP data market shares, which I think
is not correct (It should be compared to the market shares from RP data)?
Could you please kindly let me know if my syntax below is correct?
Thank you so much for your help!
Thao
|-> sample;all$
|-> reject;sprp=1$
|-> Nlogit
;lhs = cho, cset, alti
;choices = H, C, P, I, G, N
;crosstabs
;checkdata
;model:
U(H) = rl_h1 * RL_H1 + rl_h2 * RL_H2
+ fl_ * FL_H
+ cr_h1 * CR_H1 + cr_2 * CR_H2
+ lo_h1 * LO_H1
+ sa_ * SA_H
/
U(C) = com
+ rl_c1 * RL_C1 + rl_c2 * RL_C2
+ fl_ * FL_C
+ cr_c1 * CR_C1 + cr_2 * CR_C2
+ lo_c1 * LO_C1 + lo_c2 * LO_C2
+ sa_ * SA_C/
U(P) = pri
+ rl_p1 * RL_P1
+ fl_ * FL_P
+ cr_p1 * CR_P1 + cr_2 * CR_P2
+ lo_p1 * LO_P1 + lo_p2 * LO_P2
+ sa_ * SA_P/
U(I) = ind
+ rl_i1 * RL_I1 + rl_i2 * RL_I2
+ fl_ * FL_I
+ cr_i1 * CR_I1
+ lo_i1 * LO_I1
+ sa_ * SA_I/
U(G) = gov
+ rl_g1 * RL_G1
+ fl_ * FL_G
+ cr_g1 * CR_G1
+ lo_g1 * LO_G1
+ sa_ * SA_G/
U(N) = non
+ rl_n1 * RL_N1
+ fl_ * FL_N
+ cr_n1 * CR_N1
+ lo_n1 * LO_N1 + lo_n2 * LO_N2
+ sa_ * SA_N
$
+----------------------------------------------------------+
| Inspecting the data set before estimation. |
| These errors mark observations which will be skipped. |
| Row Individual = 1st row then group number of data block |
+----------------------------------------------------------+
No bad observations were found in the sample
Iterative procedure has converged
Normal exit: 5 iterations. Status=0, F= .4001834D+04
-----------------------------------------------------------------------------
Discrete choice (multinomial logit) model
Dependent variable Choice
Log likelihood function -4001.83388
Estimation based on N = 2434, K = 32
Inf.Cr.AIC = 8067.7 AIC/N = 3.315
---------------------------------------
Log likelihood R-sqrd R2Adj
ASCs only model must be fit separately
Use NLOGIT ;...;RHS=ONE$
Note: R-sqrd = 1 - logL/Logl(constants)
---------------------------------------
Chi-squared[27] = 655.81735
Prob [ chi squared > value ] = .00000
Response data are given as ind. choices
Number of obs.= 2434, skipped 0 obs
--------+--------------------------------------------------------------------
| Standard Prob. 95% Confidence
CHO| Coefficient Error z |z|>Z* Interval
--------+--------------------------------------------------------------------
RL_H1| .07005 .16175 .43 .6650 -.24697 .38707
RL_H2| .23104* .13433 1.72 .0855 -.03225 .49433
FL_| .17867*** .05585 3.20 .0014 .06920 .28813
CR_H1| .27104** .13131 2.06 .0390 .01368 .52840
CR_2| .05296 .09477 .56 .5763 -.13279 .23871
LO_H1| -.41466*** .12119 -3.42 .0006 -.65219 -.17713
SA_| .01231*** .00086 14.30 .0000 .01062 .01399
COM| -.10806 .15871 -.68 .4959 -.41913 .20300
RL_C1| .38787** .16751 2.32 .0206 .05956 .71618
RL_C2| .26364* .15216 1.73 .0832 -.03459 .56187
CR_C1| .19661 .13691 1.44 .1510 -.07174 .46496
LO_C1| -.22246* .12854 -1.73 .0835 -.47440 .02949
LO_C2| -.82061*** .15800 -5.19 .0000 -1.13027 -.51094
PRI| .40590*** .14002 2.90 .0037 .13146 .68034
RL_P1| .01672 .12447 .13 .8932 -.22725 .26068
CR_P1| .33357** .13132 2.54 .0111 .07619 .59095
LO_P1| -.95156*** .14221 -6.69 .0000 -1.23029 -.67282
LO_P2| -1.03276*** .14087 -7.33 .0000 -1.30885 -.75667
IND| -1.11685*** .15286 -7.31 .0000 -1.41645 -.81725
RL_I1| .64312*** .14665 4.39 .0000 .35570 .93054
RL_I2| .79032*** .15550 5.08 .0000 .48555 1.09508
CR_I1| .61024*** .12199 5.00 .0000 .37115 .84933
LO_I1| -.64615*** .12354 -5.23 .0000 -.88828 -.40402
GOV| -.07470 .14043 -.53 .5948 -.34993 .20053
RL_G1| -.33174** .12927 -2.57 .0103 -.58510 -.07838
CR_G1| .51684*** .12294 4.20 .0000 .27588 .75781
LO_G1| -.57783*** .12464 -4.64 .0000 -.82212 -.33354
NON| -.26651* .14521 -1.84 .0665 -.55112 .01810
RL_N1| -.08143 .13884 -.59 .5576 -.35355 .19070
CR_N1| .36292*** .13942 2.60 .0092 .08965 .63618
LO_N1| -.58506*** .15904 -3.68 .0002 -.89677 -.27336
LO_N2| -.46639*** .16171 -2.88 .0039 -.78333 -.14945
--------+--------------------------------------------------------------------
***, **, * ==> Significance at 1%, 5%, 10% level.
Model was estimated on Jul 04, 2020 at 00:17:29 PM
-----------------------------------------------------------------------------
|-> sample;all$
|-> reject;sprp=1$
|-> Nlogit
;lhs = cho, cset, alti
;choices = H, C, P, I, G, N/0.23,0.53,0.04,0.04,0.09,0.07
;checkdata
;show
'Alg=BFGS
;calibrate
; simulation
; scenario: RL_C1(c)=1/RL_C2(c)=0
;model:
U(H) = rl_h1[0.0700481] * RL_H1 + rl_h2[0.231038] * RL_H2
+ fl_[0.178665] * FL_H
+ cr_h1[0.271042] * CR_H1 + cr_2[0.0529584] * CR_H2
+ lo_h1[-0.414659] * LO_H1
+ sa_[0.0123069] * SA_H
/
U(C) = com
+ rl_c1[0.387869] * RL_C1 + rl_c2[0.263637] * RL_C2
+ fl_[0.178665] * FL_C
+ cr_c1[0.19661] * CR_C1 + cr_2[0.0529584] * CR_C2
+ lo_c1[-0.222456] * LO_C1 + lo_c2[-0.820609] * LO_C2
+ sa_[0.0123069] * SA_C
/
U(P) = pri
+ rl_p1[0.0167174] * RL_P1
+ fl_[0.178665] * FL_P
+ cr_p1[0.333571] * CR_P1 + cr_2[0.0529584] * CR_P2
+ lo_p1[-0.951557] * LO_P1 + lo_p2[-1.03276] * LO_P2
+ sa_[0.0123069] * SA_P
/
U(I) = ind
+ rl_i1[0.643119] * RL_I1 + rl_i2[0.790318] * RL_I2
+ fl_[0.178665] * FL_I
+ cr_i1[0.610239] * CR_I1
+ lo_i1[-0.64615] * LO_I1
+ sa_[0.0123069] * SA_I
/
U(G) = gov
+ rl_g1[-0.331743] * RL_G1
+ fl_[0.178665] * FL_G
+ cr_g1[0.516844] * CR_G1
+ lo_g1[-0.57783] * LO_G1
+ sa_[0.0123069] * SA_G
/
U(N) = non
+ rl_n1[-0.0814267] * RL_N1
+ fl_[0.178665] * FL_N
+ cr_n1[0.36292] * CR_N1
+ lo_n1[-0.585064] * LO_N1 + lo_n2[-0.466387] * LO_N2
+ sa_[0.0123069] * SA_N$
+----------------------------------------------------------+
| Inspecting the data set before estimation. |
| These errors mark observations which will be skipped. |
| Row Individual = 1st row then group number of data block |
+----------------------------------------------------------+
No bad observations were found in the sample
Sample proportions are marginal, not conditional.
Choices marked with * are excluded for the IIA test.
+----------------+------+
|Choice (prop.)| Count|
+----------------+------+
|H .18447| 449|
|C .17707| 431|
|P .19721| 480|
|I .17502| 426|
|G .14749| 359|
|N .11873| 289|
+----------------+------+
+---------------------------------------------+
| Discrete Choice (One Level) Model |
| Model Simulation Using Previous Estimates |
| Number of observations 2434 |
+---------------------------------------------+
+------------------------------------------------------+
|Simulations of Probability Model |
|Model: Discrete Choice (One Level) Model |
|Simulated choice set may be a subset of the choices. |
|Number of individuals is the probability times the |
|number of observations in the simulated sample. |
|Column totals may be affected by rounding error. |
|The model used was simulated with 2434 observations.|
+------------------------------------------------------+
-------------------------------------------------------------------------
Specification of scenario 1 is:
Attribute Alternatives affected Change type Value
--------- ------------------------------- ------------------- ---------
RL_C1 C Fix base at new vlu 1.000
RL_C2 C Fix base at new vlu .000
-------------------------------------------------------------------------
The simulator located 2434 observations for this scenario.
Simulated Probabilities (shares) for this scenario:
+----------+--------------+--------------+------------------+
|Choice | Base | Scenario | Scenario - Base |
| |%Share Number |%Share Number |ChgShare ChgNumber|
+----------+--------------+--------------+------------------+
|H | 18.447 449 | 17.950 437 | -.497% -12 |
|C | 17.707 431 | 19.889 484 | 2.182% 53 |
|P | 19.721 480 | 19.142 466 | -.579% -14 |
|I | 17.502 426 | 17.095 416 | -.407% -10 |
|G | 14.749 359 | 14.335 349 | -.415% -10 |
|N | 11.873 289 | 11.589 282 | -.284% -7 |
|Total |100.000 2434 |100.000 2434 | .000% 0 |
+----------+--------------+--------------+------------------+
From wgreene at stern.nyu.edu Mon Jul 6 05:22:58 2020
From: wgreene at stern.nyu.edu (William Greene)
Date: Sun, 5 Jul 2020 15:22:58 -0400
Subject: [Limdep Nlogit List] Simulation using estimates from SP data
and ASCs from RP data
In-Reply-To:
References:
Message-ID:
The same subsample,
|-> sample;all$
|-> reject;sprp=1$
is being used for both estimation and simulation.
It sounds like you want to fit the model with one subset of the
data and do the simulation with a different one. That is OK, just
set the samples accordingly for the two cases.
/B. Greene
On Fri, Jul 3, 2020 at 10:31 PM Thao Thai via Limdep <
limdep at mailman.sydney.edu.au> wrote:
> Hi Nlogit users,
>
> I am simulating some scenarios using the coefficient estimates from a CL
> model on SP data and alternative specific constants (ASCs) calibrated to
> reflect the RP data.
>
> However, the reported market shares in the base case shows the market
> shares observed in SP data. The reported market shares of the simulated
> scenarios are also compared to these SP data market shares, which I think
> is not correct (It should be compared to the market shares from RP data)?
>
> Could you please kindly let me know if my syntax below is correct?
>
> Thank you so much for your help!
> Thao
>
> |-> sample;all$
> |-> reject;sprp=1$
> |-> Nlogit
> ;lhs = cho, cset, alti
> ;choices = H, C, P, I, G, N
> ;crosstabs
> ;checkdata
> ;model:
> U(H) = rl_h1 * RL_H1 + rl_h2 * RL_H2
> + fl_ * FL_H
> + cr_h1 * CR_H1 + cr_2 * CR_H2
> + lo_h1 * LO_H1
> + sa_ * SA_H
> /
> U(C) = com
> + rl_c1 * RL_C1 + rl_c2 * RL_C2
> + fl_ * FL_C
> + cr_c1 * CR_C1 + cr_2 * CR_C2
> + lo_c1 * LO_C1 + lo_c2 * LO_C2
> + sa_ * SA_C/
> U(P) = pri
> + rl_p1 * RL_P1
> + fl_ * FL_P
> + cr_p1 * CR_P1 + cr_2 * CR_P2
> + lo_p1 * LO_P1 + lo_p2 * LO_P2
> + sa_ * SA_P/
> U(I) = ind
> + rl_i1 * RL_I1 + rl_i2 * RL_I2
> + fl_ * FL_I
> + cr_i1 * CR_I1
> + lo_i1 * LO_I1
> + sa_ * SA_I/
> U(G) = gov
> + rl_g1 * RL_G1
> + fl_ * FL_G
> + cr_g1 * CR_G1
> + lo_g1 * LO_G1
> + sa_ * SA_G/
> U(N) = non
> + rl_n1 * RL_N1
> + fl_ * FL_N
> + cr_n1 * CR_N1
> + lo_n1 * LO_N1 + lo_n2 * LO_N2
> + sa_ * SA_N
> $
> +----------------------------------------------------------+
> | Inspecting the data set before estimation. |
> | These errors mark observations which will be skipped. |
> | Row Individual = 1st row then group number of data block |
> +----------------------------------------------------------+
> No bad observations were found in the sample
>
> Iterative procedure has converged
> Normal exit: 5 iterations. Status=0, F= .4001834D+04
>
>
> -----------------------------------------------------------------------------
> Discrete choice (multinomial logit) model
> Dependent variable Choice
> Log likelihood function -4001.83388
> Estimation based on N = 2434, K = 32
> Inf.Cr.AIC = 8067.7 AIC/N = 3.315
> ---------------------------------------
> Log likelihood R-sqrd R2Adj
> ASCs only model must be fit separately
> Use NLOGIT ;...;RHS=ONE$
> Note: R-sqrd = 1 - logL/Logl(constants)
> ---------------------------------------
> Chi-squared[27] = 655.81735
> Prob [ chi squared > value ] = .00000
> Response data are given as ind. choices
> Number of obs.= 2434, skipped 0 obs
>
> --------+--------------------------------------------------------------------
> | Standard Prob. 95% Confidence
> CHO| Coefficient Error z |z|>Z* Interval
>
> --------+--------------------------------------------------------------------
> RL_H1| .07005 .16175 .43 .6650 -.24697 .38707
> RL_H2| .23104* .13433 1.72 .0855 -.03225 .49433
> FL_| .17867*** .05585 3.20 .0014 .06920 .28813
> CR_H1| .27104** .13131 2.06 .0390 .01368 .52840
> CR_2| .05296 .09477 .56 .5763 -.13279 .23871
> LO_H1| -.41466*** .12119 -3.42 .0006 -.65219 -.17713
> SA_| .01231*** .00086 14.30 .0000 .01062 .01399
> COM| -.10806 .15871 -.68 .4959 -.41913 .20300
> RL_C1| .38787** .16751 2.32 .0206 .05956 .71618
> RL_C2| .26364* .15216 1.73 .0832 -.03459 .56187
> CR_C1| .19661 .13691 1.44 .1510 -.07174 .46496
> LO_C1| -.22246* .12854 -1.73 .0835 -.47440 .02949
> LO_C2| -.82061*** .15800 -5.19 .0000 -1.13027 -.51094
> PRI| .40590*** .14002 2.90 .0037 .13146 .68034
> RL_P1| .01672 .12447 .13 .8932 -.22725 .26068
> CR_P1| .33357** .13132 2.54 .0111 .07619 .59095
> LO_P1| -.95156*** .14221 -6.69 .0000 -1.23029 -.67282
> LO_P2| -1.03276*** .14087 -7.33 .0000 -1.30885 -.75667
> IND| -1.11685*** .15286 -7.31 .0000 -1.41645 -.81725
> RL_I1| .64312*** .14665 4.39 .0000 .35570 .93054
> RL_I2| .79032*** .15550 5.08 .0000 .48555 1.09508
> CR_I1| .61024*** .12199 5.00 .0000 .37115 .84933
> LO_I1| -.64615*** .12354 -5.23 .0000 -.88828 -.40402
> GOV| -.07470 .14043 -.53 .5948 -.34993 .20053
> RL_G1| -.33174** .12927 -2.57 .0103 -.58510 -.07838
> CR_G1| .51684*** .12294 4.20 .0000 .27588 .75781
> LO_G1| -.57783*** .12464 -4.64 .0000 -.82212 -.33354
> NON| -.26651* .14521 -1.84 .0665 -.55112 .01810
> RL_N1| -.08143 .13884 -.59 .5576 -.35355 .19070
> CR_N1| .36292*** .13942 2.60 .0092 .08965 .63618
> LO_N1| -.58506*** .15904 -3.68 .0002 -.89677 -.27336
> LO_N2| -.46639*** .16171 -2.88 .0039 -.78333 -.14945
>
> --------+--------------------------------------------------------------------
> ***, **, * ==> Significance at 1%, 5%, 10% level.
> Model was estimated on Jul 04, 2020 at 00:17:29 PM
>
> -----------------------------------------------------------------------------
>
> |-> sample;all$
> |-> reject;sprp=1$
> |-> Nlogit
> ;lhs = cho, cset, alti
> ;choices = H, C, P, I, G, N/0.23,0.53,0.04,0.04,0.09,0.07
> ;checkdata
> ;show
> 'Alg=BFGS
> ;calibrate
> ; simulation
> ; scenario: RL_C1(c)=1/RL_C2(c)=0
> ;model:
> U(H) = rl_h1[0.0700481] * RL_H1 + rl_h2[0.231038] * RL_H2
> + fl_[0.178665] * FL_H
> + cr_h1[0.271042] * CR_H1 + cr_2[0.0529584] * CR_H2
> + lo_h1[-0.414659] * LO_H1
> + sa_[0.0123069] * SA_H
> /
> U(C) = com
> + rl_c1[0.387869] * RL_C1 + rl_c2[0.263637] * RL_C2
> + fl_[0.178665] * FL_C
> + cr_c1[0.19661] * CR_C1 + cr_2[0.0529584] * CR_C2
> + lo_c1[-0.222456] * LO_C1 + lo_c2[-0.820609] * LO_C2
> + sa_[0.0123069] * SA_C
> /
> U(P) = pri
> + rl_p1[0.0167174] * RL_P1
> + fl_[0.178665] * FL_P
> + cr_p1[0.333571] * CR_P1 + cr_2[0.0529584] * CR_P2
> + lo_p1[-0.951557] * LO_P1 + lo_p2[-1.03276] * LO_P2
> + sa_[0.0123069] * SA_P
> /
> U(I) = ind
> + rl_i1[0.643119] * RL_I1 + rl_i2[0.790318] * RL_I2
> + fl_[0.178665] * FL_I
> + cr_i1[0.610239] * CR_I1
> + lo_i1[-0.64615] * LO_I1
> + sa_[0.0123069] * SA_I
> /
> U(G) = gov
> + rl_g1[-0.331743] * RL_G1
> + fl_[0.178665] * FL_G
> + cr_g1[0.516844] * CR_G1
> + lo_g1[-0.57783] * LO_G1
> + sa_[0.0123069] * SA_G
> /
> U(N) = non
> + rl_n1[-0.0814267] * RL_N1
> + fl_[0.178665] * FL_N
> + cr_n1[0.36292] * CR_N1
> + lo_n1[-0.585064] * LO_N1 + lo_n2[-0.466387] * LO_N2
> + sa_[0.0123069] * SA_N$
> +----------------------------------------------------------+
> | Inspecting the data set before estimation. |
> | These errors mark observations which will be skipped. |
> | Row Individual = 1st row then group number of data block |
> +----------------------------------------------------------+
> No bad observations were found in the sample
>
>
> Sample proportions are marginal, not conditional.
> Choices marked with * are excluded for the IIA test.
> +----------------+------+
> |Choice (prop.)| Count|
> +----------------+------+
> |H .18447| 449|
> |C .17707| 431|
> |P .19721| 480|
> |I .17502| 426|
> |G .14749| 359|
> |N .11873| 289|
> +----------------+------+
>
>
> +---------------------------------------------+
> | Discrete Choice (One Level) Model |
> | Model Simulation Using Previous Estimates |
> | Number of observations 2434 |
> +---------------------------------------------+
>
> +------------------------------------------------------+
> |Simulations of Probability Model |
> |Model: Discrete Choice (One Level) Model |
> |Simulated choice set may be a subset of the choices. |
> |Number of individuals is the probability times the |
> |number of observations in the simulated sample. |
> |Column totals may be affected by rounding error. |
> |The model used was simulated with 2434 observations.|
> +------------------------------------------------------+
> -------------------------------------------------------------------------
> Specification of scenario 1 is:
> Attribute Alternatives affected Change type Value
> --------- ------------------------------- ------------------- ---------
> RL_C1 C Fix base at new vlu 1.000
> RL_C2 C Fix base at new vlu .000
> -------------------------------------------------------------------------
> The simulator located 2434 observations for this scenario.
> Simulated Probabilities (shares) for this scenario:
> +----------+--------------+--------------+------------------+
> |Choice | Base | Scenario | Scenario - Base |
> | |%Share Number |%Share Number |ChgShare ChgNumber|
> +----------+--------------+--------------+------------------+
> |H | 18.447 449 | 17.950 437 | -.497% -12 |
> |C | 17.707 431 | 19.889 484 | 2.182% 53 |
> |P | 19.721 480 | 19.142 466 | -.579% -14 |
> |I | 17.502 426 | 17.095 416 | -.407% -10 |
> |G | 14.749 359 | 14.335 349 | -.415% -10 |
> |N | 11.873 289 | 11.589 282 | -.284% -7 |
> |Total |100.000 2434 |100.000 2434 | .000% 0 |
> +----------+--------------+--------------+------------------+
> _______________________________________________
> 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/fnofCWLVXkU9jgOjF62sjW?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 john.c.whitehead at gmail.com Tue Jul 7 00:58:35 2020
From: john.c.whitehead at gmail.com (John C. Whitehead)
Date: Mon, 6 Jul 2020 10:58:35 -0400
Subject: [Limdep Nlogit List] restrictions on coefficients in mixed logit,
etc
Message-ID:
Hi all,
I'm having trouble with restrictions on coefficients in all discrete choice
models except the latent class logit.
For example, if I estimate the following model:
lclogit;Lhs=Y;Choices=yes,no;Rhs=x,y,z;pds=3;lcm;pts=2;
rst=b1,b2,b3,0,0,0$
NLogit will constrain the coefficients in the second class to zero.
But, if I estimate the scaled model:
smnlogit;Lhs=Y;Choices=yes,no;Rhs=x,y,z;pds=3;halton;pts=25;
rst=0,b2,b3$
It produces output as if the restriction was not given (i.e., it estimates
a number for b1).
I've searched the manual and it says restrictions should work in these
models (unless I've missed something, I haven't read every word).
Any help would be appreciated!
Thanks,
John Whitehead
From avassilopoulos.aua at gmail.com Tue Jul 7 20:15:16 2020
From: avassilopoulos.aua at gmail.com (Achilleas' Gmail)
Date: Tue, 7 Jul 2020 13:15:16 +0300
Subject: [Limdep Nlogit List] restrictions on coefficients in mixed
logit, etc
In-Reply-To:
References:
Message-ID: <000001d65447$844f77a0$8cee66e0$@gmail.com>
Hi, I've noticed the same.
Neither "; CML:" seems to work with smnlogit.
Best,
Achilleas Vassilopoulos
-----Original Message-----
From: Limdep On Behalf Of John C.
Whitehead via Limdep
Sent: Monday, July 6, 2020 17:59
To: Limdep and Nlogit Mailing List
Cc: John C. Whitehead
Subject: [Limdep Nlogit List] restrictions on coefficients in mixed logit,
etc
Hi all,
I'm having trouble with restrictions on coefficients in all discrete choice
models except the latent class logit.
For example, if I estimate the following model:
lclogit;Lhs=Y;Choices=yes,no;Rhs=x,y,z;pds=3;lcm;pts=2;
rst=b1,b2,b3,0,0,0$
NLogit will constrain the coefficients in the second class to zero.
But, if I estimate the scaled model:
smnlogit;Lhs=Y;Choices=yes,no;Rhs=x,y,z;pds=3;halton;pts=25;
rst=0,b2,b3$
It produces output as if the restriction was not given (i.e., it estimates a
number for b1).
I've searched the manual and it says restrictions should work in these
models (unless I've missed something, I haven't read every word).
Any help would be appreciated!
Thanks,
John Whitehead
_______________________________________________
Limdep site list
Limdep at mailman.sydney.edu.au
http://limdep.itls.usyd.edu.au
From avassilopoulos.aua at gmail.com Tue Jul 7 22:30:20 2020
From: avassilopoulos.aua at gmail.com (Achilleas' Gmail)
Date: Tue, 7 Jul 2020 15:30:20 +0300
Subject: [Limdep Nlogit List] restrictions on coefficients in mixed
logit, etc
References:
Message-ID: <000801d6545a$61d754a0$2585fde0$@gmail.com>
Hello again,
Coming back to this issue, "; Fix = name[0] " seems to be working well with
smnlogit.
namelist ; x = gc,ttme,invc,invt,one $
smnlogit ; Lhs = Mode ; Choices = air,train,bus,car
; Rhs = x
; Fix = ttme[0] $
Best,
Achilleas Vassilopoulos
-----Original Message-----
From: Achilleas' Gmail
Sent: Tuesday, July 7, 2020 13:15
To: 'Limdep and Nlogit Mailing List'
Subject: RE: [Limdep Nlogit List] restrictions on coefficients in mixed
logit, etc
Hi, I've noticed the same.
Neither "; CML:" seems to work with smnlogit.
Best,
Achilleas Vassilopoulos
-----Original Message-----
From: Limdep On Behalf Of John C.
Whitehead via Limdep
Sent: Monday, July 6, 2020 17:59
To: Limdep and Nlogit Mailing List
Cc: John C. Whitehead
Subject: [Limdep Nlogit List] restrictions on coefficients in mixed logit,
etc
Hi all,
I'm having trouble with restrictions on coefficients in all discrete choice
models except the latent class logit.
For example, if I estimate the following model:
lclogit;Lhs=Y;Choices=yes,no;Rhs=x,y,z;pds=3;lcm;pts=2;
rst=b1,b2,b3,0,0,0$
NLogit will constrain the coefficients in the second class to zero.
But, if I estimate the scaled model:
smnlogit;Lhs=Y;Choices=yes,no;Rhs=x,y,z;pds=3;halton;pts=25;
rst=0,b2,b3$
It produces output as if the restriction was not given (i.e., it estimates a
number for b1).
I've searched the manual and it says restrictions should work in these
models (unless I've missed something, I haven't read every word).
Any help would be appreciated!
Thanks,
John Whitehead
_______________________________________________
Limdep site list
Limdep at mailman.sydney.edu.au
http://limdep.itls.usyd.edu.au
From john.c.whitehead at gmail.com Wed Jul 8 10:34:42 2020
From: john.c.whitehead at gmail.com (John C. Whitehead)
Date: Tue, 7 Jul 2020 20:34:42 -0400
Subject: [Limdep Nlogit List] restrictions on coefficients in mixed
logit, etc
In-Reply-To: <000801d6545a$61d754a0$2585fde0$@gmail.com>
References:
<000801d6545a$61d754a0$2585fde0$@gmail.com>
Message-ID:
;Fix = x[0]$ is working for me with all models!
But, I'd still like to constrain some parameters to be equal ...
On Tue, Jul 7, 2020 at 8:30 AM Achilleas' Gmail via Limdep <
limdep at mailman.sydney.edu.au> wrote:
> Hello again,
>
> Coming back to this issue, "; Fix = name[0] " seems to be working well
> with
> smnlogit.
>
> namelist ; x = gc,ttme,invc,invt,one $
> smnlogit ; Lhs = Mode ; Choices = air,train,bus,car
> ; Rhs = x
> ; Fix = ttme[0] $
>
> Best,
> Achilleas Vassilopoulos
>
> -----Original Message-----
> From: Achilleas' Gmail
> Sent: Tuesday, July 7, 2020 13:15
> To: 'Limdep and Nlogit Mailing List'
> Subject: RE: [Limdep Nlogit List] restrictions on coefficients in mixed
> logit, etc
>
> Hi, I've noticed the same.
>
> Neither "; CML:" seems to work with smnlogit.
>
> Best,
> Achilleas Vassilopoulos
>
> -----Original Message-----
> From: Limdep On Behalf Of John C.
> Whitehead via Limdep
> Sent: Monday, July 6, 2020 17:59
> To: Limdep and Nlogit Mailing List
> Cc: John C. Whitehead
> Subject: [Limdep Nlogit List] restrictions on coefficients in mixed logit,
> etc
>
> Hi all,
>
> I'm having trouble with restrictions on coefficients in all discrete choice
> models except the latent class logit.
>
> For example, if I estimate the following model:
>
> lclogit;Lhs=Y;Choices=yes,no;Rhs=x,y,z;pds=3;lcm;pts=2;
> rst=b1,b2,b3,0,0,0$
>
> NLogit will constrain the coefficients in the second class to zero.
>
> But, if I estimate the scaled model:
>
> smnlogit;Lhs=Y;Choices=yes,no;Rhs=x,y,z;pds=3;halton;pts=25;
> rst=0,b2,b3$
>
> It produces output as if the restriction was not given (i.e., it estimates
> a
> number for b1).
>
> I've searched the manual and it says restrictions should work in these
> models (unless I've missed something, I haven't read every word).
>
> Any help would be appreciated!
>
> Thanks,
>
> John Whitehead
> _______________________________________________
> Limdep site list
> Limdep at mailman.sydney.edu.au
> http://limdep.itls.usyd.edu.au
>
>
> _______________________________________________
> Limdep site list
> Limdep at mailman.sydney.edu.au
> http://limdep.itls.usyd.edu.au
>
>
From david.hensher at sydney.edu.au Wed Jul 8 10:36:58 2020
From: david.hensher at sydney.edu.au (David Hensher)
Date: Wed, 08 Jul 2020 10:36:58 +1000
Subject: [Limdep Nlogit List] restrictions on coefficients in mixed
logit, etc
In-Reply-To:
References: <000801d6545a$61d754a0$2585fde0$@gmail.com>
Message-ID: <5F05152A.3000306@sydney.edu.au>
To make betas equal give them the same names
David
On 8/07/2020 10:34 AM, John C. Whitehead via Limdep wrote:
> ;Fix = x[0]$ is working for me with all models!
>
> But, I'd still like to constrain some parameters to be equal ...
>
> On Tue, Jul 7, 2020 at 8:30 AM Achilleas' Gmail via Limdep<
> limdep at mailman.sydney.edu.au> wrote:
>
>
>> Hello again,
>>
>> Coming back to this issue, "; Fix = name[0] " seems to be working well
>> with
>> smnlogit.
>>
>> namelist ; x = gc,ttme,invc,invt,one $
>> smnlogit ; Lhs = Mode ; Choices = air,train,bus,car
>> ; Rhs = x
>> ; Fix = ttme[0] $
>>
>> Best,
>> Achilleas Vassilopoulos
>>
>> -----Original Message-----
>> From: Achilleas' Gmail
>> Sent: Tuesday, July 7, 2020 13:15
>> To: 'Limdep and Nlogit Mailing List'
>> Subject: RE: [Limdep Nlogit List] restrictions on coefficients in mixed
>> logit, etc
>>
>> Hi, I've noticed the same.
>>
>> Neither "; CML:" seems to work with smnlogit.
>>
>> Best,
>> Achilleas Vassilopoulos
>>
>> -----Original Message-----
>> From: Limdep On Behalf Of John C.
>> Whitehead via Limdep
>> Sent: Monday, July 6, 2020 17:59
>> To: Limdep and Nlogit Mailing List
>> Cc: John C. Whitehead
>> Subject: [Limdep Nlogit List] restrictions on coefficients in mixed logit,
>> etc
>>
>> Hi all,
>>
>> I'm having trouble with restrictions on coefficients in all discrete choice
>> models except the latent class logit.
>>
>> For example, if I estimate the following model:
>>
>> lclogit;Lhs=Y;Choices=yes,no;Rhs=x,y,z;pds=3;lcm;pts=2;
>> rst=b1,b2,b3,0,0,0$
>>
>> NLogit will constrain the coefficients in the second class to zero.
>>
>> But, if I estimate the scaled model:
>>
>> smnlogit;Lhs=Y;Choices=yes,no;Rhs=x,y,z;pds=3;halton;pts=25;
>> rst=0,b2,b3$
>>
>> It produces output as if the restriction was not given (i.e., it estimates
>> a
>> number for b1).
>>
>> I've searched the manual and it says restrictions should work in these
>> models (unless I've missed something, I haven't read every word).
>>
>> Any help would be appreciated!
>>
>> Thanks,
>>
>> John Whitehead
>> _______________________________________________
>> Limdep site list
>> Limdep at mailman.sydney.edu.au
>> http://limdep.itls.usyd.edu.au
>>
>>
>> _______________________________________________
>> Limdep site list
>> Limdep at mailman.sydney.edu.au
>> http://limdep.itls.usyd.edu.au
>>
>>
>>
> _______________________________________________
> Limdep site list
> Limdep at mailman.sydney.edu.au
> http://limdep.itls.usyd.edu.au
>
>
--
DAVID HENSHER FASSA, PhD| Professor and Founding Director
Institute of Transport and Logistics Studies | The University of Sydney Business School
THE UNIVERSITY OF SYDNEY
Rm 201, Building H73| The University of Sydney | NSW | 2006
Street Address: 378 Abercrombie St, Darlington NSW 2008
T +61 2 9114 1871 | F +61 2 9114 1863 | M +61 418 433 057
E David.Hensher at sydney.edu.au | W sydney.edu.au/business/itls |W http://sydney.edu.au/business/itls/staff/davidh
Celebrating 30 years of ITLS: 1991-2020 https://protect-au.mimecast.com/s/agrmCNLJyQUloMzJIm66Ac?domain=youtu.be
ERA Rank 5 (Transportation and Freight Services)
Co-Founder of the International Conference Series on Competition and Ownership of Land Passenger Transport (The 'Thredbo' Series)
https://protect-au.mimecast.com/s/b4kNCOMKzVTR32MlHvBdnQ?domain=thredbo-conference-series.org
https://www.linkedin.com/company/28450714
Second edition of Applied Choice Analysis now available at https://protect-au.mimecast.com/s/ebYYCP7LAXfGQ5P9U1C9AY?domain=cambridge.org
Nlogit is the most popular software for choice modellers. See https://protect-au.mimecast.com/s/hEYDCQnMBZfA7NYgtkMmi2?domain=limdep.com
CRICOS 00026A
This email plus any attachments to it are confidential. Any unauthorised use is strictly prohibited. If you receive this email in error, please delete it and any attachments.
Please think of our environment and only print this e-mail if necessary.
From sdsuh at hotmail.com Sat Jul 18 13:41:07 2020
From: sdsuh at hotmail.com (Daniel Suh)
Date: Sat, 18 Jul 2020 03:41:07 +0000
Subject: [Limdep Nlogit List] NLOGIT 4, dialog box language issue
Message-ID:
Hello,
I am having dialog boxes in Japanese language whenever I run NLOGIT (version 4). It is very peculiar since all NLOGIT program menus are in English. I do not have Japanese in window(windows 10, 64 bit) language settings.
I remember I did have the same issue some year back, and somehow I did resolve the issue. Problem is I do not remember what I did that time, and all the google searches now do not give me any clue.
Could you somebody help me with this??
Thanks.
From florian.neubauer at uconn.edu Wed Jul 29 01:15:14 2020
From: florian.neubauer at uconn.edu (Florian Neubauer)
Date: Tue, 28 Jul 2020 11:15:14 -0400
Subject: [Limdep Nlogit List] Export to Excel
Message-ID:
Hi all,
I am completely new to Limdep/Nlogit. I?m estimating a stochastic frontier model and want to export the results to an Excel file instead of doing everything manually. Is that possible and if so, could you send me an example of the code?
Thanks,
Florian
From avassilopoulos.aua at gmail.com Wed Jul 29 20:01:48 2020
From: avassilopoulos.aua at gmail.com (Achilleas' Gmail)
Date: Wed, 29 Jul 2020 13:01:48 +0300
Subject: [Limdep Nlogit List] Export to Excel
In-Reply-To:
References:
Message-ID: <003301d6658f$46f88870$d4e99950$@gmail.com>
Hi Florian,
Try the following:
OPEN ; Export = "YOUR_PATH\RESULTS.csv" $
FRONTIER ; Lhs = y ; Rhs = x ; Export $
CLOSE ; Export $
Best,
Achilleas
-----Original Message-----
From: Limdep On Behalf Of Florian Neubauer
Sent: Tuesday, July 28, 2020 18:15
To: limdep at mailman.sydney.edu.au
Subject: [Limdep Nlogit List] Export to Excel
Hi all,
I am completely new to Limdep/Nlogit. I?m estimating a stochastic frontier model and want to export the results to an Excel file instead of doing everything manually. Is that possible and if so, could you send me an example of the code?
Thanks,
Florian
_______________________________________________
Limdep site list
Limdep at mailman.sydney.edu.au
http://limdep.itls.usyd.edu.au