Railway station choice predictor
Railway station choice predictor
Station choice predictions
Catchment maps only include postcodes where the station probability is >= 5%
All models were calibrated using combined LATIS and WG datasets.
Model TE19
-----------------------------------------------------------------------------
Discrete choice (multinomial logit) model
Dependent variable Choice
Log likelihood function -9650.85340
Estimation based on N = 14422, K = 8
Inf.Cr.AIC = 19317.7 AIC/N = 1.339
R2Adj = 0.71
---------------------------------------
Response data are given as ind. choices
Number of obs.= 14422, skipped 0 obs
--------+--------------------------------------------------------------------
| Standard Prob. 95% Confidence
CHOICE| Coefficient Error z |z|>Z* Interval
--------+--------------------------------------------------------------------
NEAREST| .69065*** .03744 18.44 .0000 .61726 .76404
SQR_DIST| -2.26183*** .04016 -56.31 .0000 -2.34056 -2.18311
CAT_F| -.67672*** .04226 -16.01 .0000 -.75954 -.59390
LN_DFQAL| 1.19857*** .03468 34.57 .0000 1.13061 1.26654
CCTV| 1.07082*** .12464 8.59 .0000 .82652 1.31512
CPSPACES| .00132*** .7988D-04 16.48 .0000 .00116 .00147
TICKETM| .98392*** .05156 19.08 .0000 .88286 1.08497
BUSES| .75848*** .05574 13.61 .0000 .64924 .86773
--------+--------------------------------------------------------------------
nnnnn.D-xx or D+xx => multiply by 10 to -xx or +xx.
***, **, * ==> Significance at 1%, 5%, 10% level.
Model was estimated on Aug 30, 2017 at 02:49:49 PM
-----------------------------------------------------------------------------
Model TE24 - including accessibility term
-----------------------------------------------------------------------------
Discrete choice (multinomial logit) model
Dependent variable Choice
Log likelihood function -9645.65795
Estimation based on N = 14422, K = 9
Inf.Cr.AIC = 19309.3 AIC/N = 1.339
R2Adj = 0.71
---------------------------------------
Response data are given as ind. choices
Number of obs.= 14422, skipped 0 obs
--------+--------------------------------------------------------------------
| Standard Prob. 95% Confidence
CHOICE| Coefficient Error z |z|>Z* Interval
--------+--------------------------------------------------------------------
NEAREST| .69077*** .03752 18.41 .0000 .61723 .76431
SQR_DIST| -2.26517*** .04017 -56.39 .0000 -2.34390 -2.18643
CAT_F| -.63827*** .04392 -14.53 .0000 -.72435 -.55219
LN_DFQAL| 1.21447*** .03507 34.63 .0000 1.14574 1.28319
CCTV| 1.07603*** .12471 8.63 .0000 .83161 1.32046
CPSPACES| .00120*** .8783D-04 13.67 .0000 .00103 .00137
TICKETM| .96331*** .05187 18.57 .0000 .86164 1.06497
BUSES| .73079*** .05642 12.95 .0000 .62020 .84137
LN_FNMWA| -.14129*** .04388 -3.22 .0013 -.22729 -.05530
--------+--------------------------------------------------------------------
nnnnn.D-xx or D+xx => multiply by 10 to -xx or +xx.
***, **, * ==> Significance at 1%, 5%, 10% level.
Model was estimated on Aug 30, 2017 at 02:53:28 PM
-----------------------------------------------------------------------------
TE12 - nearest main added to choice set
-----------------------------------------------------------------------------
Discrete choice (multinomial logit) model
Dependent variable Choice
Log likelihood function -11707.20038
Estimation based on N = 15047, K = 8
Inf.Cr.AIC = 23430.4 AIC/N = 1.557
R2Adj = 0.67
---------------------------------------
Response data are given as ind. choices
Number of obs.= 15047, skipped 0 obs
--------+--------------------------------------------------------------------
| Standard Prob. 95% Confidence
CHOICE| Coefficient Error z |z|>Z* Interval
--------+--------------------------------------------------------------------
NEAREST| .97183*** .03292 29.52 .0000 .90730 1.03635
SQR_DIST| -1.83562*** .02822 -65.04 .0000 -1.89093 -1.78030
CAT_F| -.58165*** .03929 -14.80 .0000 -.65866 -.50464
LN_DFQAL| 1.33186*** .03043 43.77 .0000 1.27222 1.39149
CCTV| .85563*** .11432 7.48 .0000 .63157 1.07968
CPSPACES| .00099*** .7377D-04 13.40 .0000 .00084 .00113
TICKETM| .97142*** .04715 20.60 .0000 .87902 1.06383
BUSES| .61587*** .05187 11.87 .0000 .51420 .71754
--------+--------------------------------------------------------------------
nnnnn.D-xx or D+xx => multiply by 10 to -xx or +xx.
***, **, * ==> Significance at 1%, 5%, 10% level.
Model was estimated on Aug 30, 2017 at 04:03:32 PM
-----------------------------------------------------------------------------