Table of Contents

Making survey data monotonic

As one can observe in plots of the survey indices (see top here), these are frequently not monotonic. For this reason we make the survey indices nonicreasing by removing the offending indides, and repeat the experiments in the section about choosing the number of cohorts per run. To be more specific, we remove the observations of a cohort which are smaller than subsequent observations.

Summary

TBD.

First cohort born in 1994

Num. cohorts Trajectories Estimates
1
cod.par
# Number of parameters = 5  Objective function value = -3.08697  Maximum gradient component = 1.69126e-05
# N0:
 3.29757
# q02:
 0.605650
# q3plus:
 0.354787
# logs:
-0.885871081891
# M:
0.499998979520
 
2
cod.par
# Number of parameters = 6  Objective function value = -6.85054  Maximum gradient component = 6.41162e-05
# N0:
 3.77137 6.68468
# q02:
 0.412709
# q3plus:
 0.290194
# logs:
-0.928159057590
# M:
0.499999742594
 
3
cod.par
# Number of parameters = 7  Objective function value = -10.3721  Maximum gradient component = 2.04829e-05
# N0:
 5.24794 8.95292 9.36205
# q02:
 0.275668
# q3plus:
 0.183000
# logs:
-0.932172674583
# M:
0.499999872250
 
4
cod.par
 # Number of parameters = 8  Objective function value = -9.88239  Maximum gradient component = 1.61960e-05
# N0:
 5.34875 9.10636 9.49745 9.18773
# q02:
 0.261406
# q3plus:
 0.180118
# logs:
-0.808824614075
# M:
0.499999804090

First cohort born in 1995

Num. cohorts Trajectories Estimates
1
cod.par
# Number of parameters = 5  Objective function value = -5.15249  Maximum gradient component = 2.53430e-06
# N0:
 7.71856
# q02:
 0.278557
# q3plus:
 0.237080
# logs:
-1.14406139971
# M:
0.499999698632
 
2
cod.par
# Number of parameters = 6  Objective function value = -9.86903  Maximum gradient component = 6.77050e-05
# N0:
 11.6958 11.8592
# q02:
 0.182385
# q3plus:
 0.132361
# logs:
-1.11681438596
# M:
0.499999817025
 
3
cod.par
# Number of parameters = 7  Objective function value = -8.59159  Maximum gradient component = 3.32696e-05
# N0:
 10.8458 11.0748 10.9318
# q02:
 0.196332
# q3plus:
 0.145901
# logs:
-0.857982907239
# M:
0.499999584145
 
4
cod.par
# Number of parameters = 8  Objective function value = -9.29134  Maximum gradient component = 3.16538e-05
# N0:
 21.7913 21.2949 22.0305 14.3428
# q02:
 0.0797296
# q3plus:
 0.0678989
# logs:
-0.790354539022
# M:
0.499999844255
 

First cohort born in 1996

Num. cohorts Trajectories Estimates
1
cod.par
# Number of parameters = 5  Objective function value = -5.44246  Maximum gradient component = 8.80377e-05
# N0:
 21.7166
# q02:
 0.0988148
# q3plus:
 0.0620195
# logs:
-1.18030733268
# M:
0.499998776989
 
2
cod.par
# Number of parameters = 6  Objective function value = -4.36038  Maximum gradient component = 6.45188e-06
# N0:
 13.3269 13.3847
# q02:
 0.160961
# q3plus:
 0.113024
# logs:
-0.772523846511
# M:
0.499978340306
 
3
cod.par
# Number of parameters = 7  Objective function value = -5.97017  Maximum gradient component = 4.44568e-05
# N0:
 28.4555 29.7430 18.6543
# q02:
 0.0558861
# q3plus:
 0.0500025
# logs:
-0.748757059934
# M:
0.499999418221
 
4
cod.par
# Number of parameters = 8  Objective function value = -7.85681  Maximum gradient component = 4.30842e-05
# N0:
 27.0583 28.4144 17.6760 11.9686
# q02:
 0.0500004
# q3plus:
 0.0500001
# logs:
-0.753445416737
# M:
0.486292217940
 

First cohort born in 1997

Num. cohorts Trajectories Estimates
1
cod.par
# Number of parameters = 5  Objective function value = -0.576982  Maximum gradient component = 6.53503e-06
# N0:
 3.18481
# q02:
 0.572871
# q3plus:
 0.333259
# logs:
-0.572122768699
# M:
0.389392591051
 
2
cod.par
# Number of parameters = 6  Objective function value = -2.56523  Maximum gradient component = 2.18165e-05
# N0:
 28.7168 18.0327
# q02:
 0.0500009
# q3plus:
 0.0543061
# logs:
-0.660327137031
# M:
0.499995000270
 
3
cod.par
# Number of parameters = 7  Objective function value = -5.07237  Maximum gradient component = 8.93392e-05
# N0:
 23.7926 14.7673 9.86596
# q02:
 0.0500001
# q3plus:
 0.0544865
# logs:
-0.720537843156
# M:
0.464181122587
 
4
cod.par
# Number of parameters = 8  Objective function value = -8.16818  Maximum gradient component = 4.40661e-06
# N0:
 21.6570 13.2731 8.74892 18.7506
# q02:
 0.0517102
# q3plus:
 0.0500001
# logs:
-0.772272656988
# M:
0.435222747275