NZRSE conf

Extendable projection of social contact matrices

Nicholas Tierney

Telethon Kids Institute

The story so far

• 2008-2012: Undergraduate + honours in Psychology

• 2013 - 2017: PhD Statistics, QUT

  • Exploratory Data Analysis (EDA)
  • Bayesian / Geospatial statistics / Optimal placement

• 2018 - 2020: Research Fellow / Lecturer at Monash

  • Design and improve tools for EDA

• 2020 - 2022: Research Software Engineer @ Telethon Kids Institute

  • Maintain and design tools for data analysis

visdat::vis_dat(airquality)

naniar::gg_miss_upset(riskfactors)

brolgar - take spaghetti

brolgar - spread spaghetti

Infectious Disease Modelling

• I was briefly part of a team advising Australian Government for COVID response in 2021

• Diseases like COVID19 and Influenza spread through face to face social contact

• I think we’re all familiar enough with COVID to know what I mean here.

• Describe which 3 people had contact:

  • James has had contact with Luke

  • Nick hasn’t had contact with either

So say we have 3 people and we want to describe who have had face to face contact in the past week. Let’s take myself and my two brothers. They’re both in Brisbane and I’m not, so I haven’t seen them. We might say:

Example: visually

This shows whether people had contact, yes or no. We can see that everyone had contact with one another except for Nick, and with Nick. Note that we could change this from something logical - “did they have contact” to “how many times did you have contact?”

Example: matrix

      James  Luke  Nick
James  TRUE  TRUE FALSE
Luke   TRUE  TRUE FALSE
Nick  FALSE FALSE  TRUE

Logical –> Numeric

So now we have not just, “Did you have contact?”, but, “How many times did you have contact?” Instead of individual people, we could change this to age groups.

People –> Age groups

What do you do with this?

• Well, if we know how many times people have contact, then we can have an idea of which age groups will spread COVID

• Simulate how many COVID cases would get transmitted

• Explore how vaccination reduces transmission

• Do this for different areas in Australia

Do we know how much contact people have?

• We don’t. Well, not in Australia. Yet.

• You need to conduct a surveys where people diary the amount and manner of daily contacts they have.

• Mossong et al have this for 8 countries in Europe

• So we have a pretty good idea of the amount of contact people have in Europe.

Method of Mossong et al

• Random weekday: record every person in contact with.
• Physical contact:
  • skin-to-skin contact (kiss or handshake)
• Non-physical contact:
  • two-way conversation with 3+ words in physical presence of a person

Method of Mossong et al

• Participants provide info on:
  • age and sex of each contact person
  • location (home, work, school, leisure, transport, or other)
  • time spent together
  • frequency of usual contacts with this individual

Contact surveys are 💰 💰 💰

• Part of why we don’t have them in Australia
• Can we get a contact matrix for a country not on the list?
• We need to have a best guess from existing surveys

Synthetic contact matrices

• Pre-existing statistical methodologies project empirical contact matrices to new countries.
• New projected contact matrices: “synthetic contact matrices”
• Use existing contact surveys (e.g., Mossong)
• Use setting-specific survey data on household, school, classroom, and workplace composition
• Combine with empirical data on contact patterns in Europe using statistical models

Prem et al

• Prem et al. have one of the most widely used approaches to synthetic contact matrices
• Extensively applied across infectious diseases applications
• Initially provided 155 matrices for 155 countries (177 later in 2020)

Brief explainer of Prem

• There is a lot more to the method! Not enough time.

• Core ideas:

  • Build a model from the POLYMOD study, predicting number of contacts
  • Incorporate key age information for different locations (home, work, school, other)
  • Extrapolate the number of contacts to different countries, using age information from those countries
  • Create a new contact matrix for a given country

image of spreadsheet booklet of contact matrices

The main issue with this is that we don’t have much control over the outputs.

And maybe that doesn’t matter to you if you just need to look at the contact matrices for a given country.

But what we needed to do is get contact matrices for a given age population structure.

Populations are different

Populations are different

What we want

• We want to be able use an age distribution population

# A tibble: 18 × 4
   lga           lower.age.limit  year population
   <chr>                   <dbl> <dbl>      <dbl>
 1 Fairfield (C)               0  2020      12261
 2 Fairfield (C)               5  2020      13093
 3 Fairfield (C)              10  2020      13602
 4 Fairfield (C)              15  2020      14323
 5 Fairfield (C)              20  2020      15932
 6 Fairfield (C)              25  2020      16190
 7 Fairfield (C)              30  2020      14134
 8 Fairfield (C)              35  2020      13034
 9 Fairfield (C)              40  2020      12217
10 Fairfield (C)              45  2020      13449
11 Fairfield (C)              50  2020      13419
12 Fairfield (C)              55  2020      13652
13 Fairfield (C)              60  2020      12907
14 Fairfield (C)              65  2020      10541
15 Fairfield (C)              70  2020       8227
16 Fairfield (C)              75  2020       5598
17 Fairfield (C)              80  2020       4006
18 Fairfield (C)              85  2020       4240

What we want

• Input age distribution population and get out a contact matrix.

extrapolate_from_data(
  population = fairfield
)

How do we extend this method?

• Code is provided, but a few key issues:

  • It was code not written for reuse (code vs software)

    • No clear interface on how to get inputs for a given country or region.

    • Challenging to see which bits of code matched which methods

Building our own extension

• Nick Golding wrote a new method that was able to be more flexible, using GAMs instead of Bayesian approach.

• I was tasked with writing software from initial model fitting code, with guidance from Nick

• Named the package, conmat, (repo), creating a home for code for others to contribute to.

• In R package form, this made it easier for us to develop and extend software on demand for our needs

Demonstration of conmat

library(conmat)
fairfield <- abs_age_lga("Fairfield (C)")
fairfield

# A tibble: 18 × 4
   lga           lower.age.limit  year population
   <chr>                   <dbl> <dbl>      <dbl>
 1 Fairfield (C)               0  2020      12261
 2 Fairfield (C)               5  2020      13093
 3 Fairfield (C)              10  2020      13602
 4 Fairfield (C)              15  2020      14323
 5 Fairfield (C)              20  2020      15932
 6 Fairfield (C)              25  2020      16190
 7 Fairfield (C)              30  2020      14134
 8 Fairfield (C)              35  2020      13034
 9 Fairfield (C)              40  2020      12217
10 Fairfield (C)              45  2020      13449
11 Fairfield (C)              50  2020      13419
12 Fairfield (C)              55  2020      13652
13 Fairfield (C)              60  2020      12907
14 Fairfield (C)              65  2020      10541
15 Fairfield (C)              70  2020       8227
16 Fairfield (C)              75  2020       5598
17 Fairfield (C)              80  2020       4006
18 Fairfield (C)              85  2020       4240

Extrapolate to a new population

synthetic_fairfield <- extrapolate_polymod(
 population = fairfield
)
synthetic_fairfield

$home
              [0,5)     [5,10)    [10,15)    [15,20)    [20,25)    [25,30)
[0,5)    0.55937722 0.38605578 0.17917034 0.11728346 0.17494283 0.37757782
[5,10)   0.45004867 0.74634125 0.45633287 0.17768035 0.11763396 0.20425143
[10,15)  0.27441817 0.48678530 0.83400880 0.44036841 0.16102946 0.11745813
[15,20)  0.22855433 0.22835607 0.44905886 0.76940487 0.38040014 0.14104260
[20,25)  0.33238719 0.17819852 0.18860264 0.37959322 0.70468036 0.35514830
[25,30)  0.56775631 0.26871520 0.14213635 0.14225059 0.33082841 0.67959783
[30,35)  0.66274730 0.47809697 0.22250688 0.10260798 0.11221744 0.30171394
[35,40)  0.44730888 0.57299240 0.41853198 0.16576020 0.07660462 0.09834327
[40,45)  0.22940356 0.38619579 0.52331490 0.33408950 0.12806613 0.06637604
[45,50)  0.15650537 0.19539202 0.35968238 0.44381733 0.27749613 0.11513026
[50,55)  0.16636819 0.12546892 0.16872119 0.29621590 0.36490096 0.24243058
[55,60)  0.18849101 0.12613868 0.09431281 0.12533086 0.22642390 0.29336326
[60,65)  0.15802431 0.13593426 0.08217401 0.05998173 0.08471409 0.15999501
[65,70)  0.09570148 0.10835705 0.07957739 0.04579578 0.03616018 0.05388495
[70,75)  0.05294941 0.06406294 0.06090246 0.04229044 0.02719440 0.02414816
[75,Inf) 0.05027334 0.06177428 0.06258821 0.05725974 0.05241443 0.04957133
            [30,35)    [35,40)    [40,45)    [45,50)    [50,55)    [55,60)
[0,5)    0.52917112 0.35928119 0.16939406 0.11519344 0.13626562 0.17966074
[5,10)   0.48329112 0.65571410 0.41431360 0.19359920 0.13964318 0.17105926
[10,15)  0.23677472 0.54327687 0.65979878 0.38734840 0.18766575 0.14156558
[15,20)  0.11969489 0.24277750 0.49308926 0.53180522 0.30843042 0.15756021
[20,25)  0.14170296 0.12618114 0.23349604 0.41373381 0.42158607 0.25650758
[25,30)  0.34032571 0.14130462 0.12607919 0.21020383 0.33972088 0.35195852
[30,35)  0.59870582 0.29219734 0.13057778 0.11693882 0.17967313 0.28378112
[35,40)  0.26869822 0.47143003 0.24209121 0.12209763 0.10846906 0.15915326
[40,45)  0.09510824 0.22856149 0.37957646 0.22772625 0.12424619 0.10744873
[45,50)  0.06897799 0.09490577 0.20492757 0.37893998 0.25668045 0.14088809
[50,55)  0.11465273 0.07258549 0.09102565 0.20497849 0.43964817 0.30555261
[55,60)  0.21618480 0.11053156 0.06773859 0.08440917 0.21854557 0.50930064
[60,65)  0.22453226 0.17614479 0.08929441 0.05439145 0.07522306 0.22078229
[65,70)  0.10788631 0.15884264 0.12268457 0.06109755 0.04006227 0.06426870
[70,75)  0.03740429 0.07617792 0.10953296 0.08157081 0.04217937 0.03192569
[75,Inf) 0.04806763 0.05451776 0.08469951 0.12256805 0.12477643 0.10136536
            [60,65)    [65,70)    [70,75)  [75,Inf)
[0,5)    0.17300920 0.12313121 0.09835800 0.1303466
[5,10)   0.20946349 0.17823707 0.12823699 0.1623448
[10,15)  0.16474834 0.18129131 0.15561482 0.1679962
[15,20)  0.11792451 0.12728375 0.14073654 0.1589760
[20,25)  0.13472693 0.09620881 0.10322754 0.1520150
[25,30)  0.22299066 0.11727108 0.08591414 0.1419087
[30,35)  0.29909854 0.19199785 0.10800225 0.1272426
[35,40)  0.24610389 0.25556251 0.17457113 0.1259902
[40,45)  0.14895519 0.21688555 0.23123415 0.1615744
[45,50)  0.11429749 0.14456195 0.20629207 0.2292049
[50,55)  0.15763916 0.11714872 0.14181890 0.2705644
[55,60)  0.33064807 0.15801247 0.11408222 0.2478890
[60,65)  0.49089454 0.29607266 0.14276766 0.1795526
[65,70)  0.19216238 0.39381011 0.23939574 0.1323827
[70,75)  0.05648987 0.15973163 0.30328341 0.1473635
[75,Inf) 0.08199238 0.09569276 0.19486913 0.4023698

$work
                [0,5)       [5,10)     [10,15)     [15,20)    [20,25)
[0,5)    0.0094683883 0.0032734722 0.002816215 0.011884040 0.04908195
[5,10)   0.0045967037 0.0060609400 0.005933911 0.017288984 0.06254511
[10,15)  0.0019812806 0.0036442034 0.017193699 0.043289867 0.10052098
[15,20)  0.0015116296 0.0021818055 0.015204953 0.139122225 0.23635155
[20,25)  0.0019613629 0.0022899930 0.012302327 0.137345715 0.59402769
[25,30)  0.0027662748 0.0024147272 0.009761121 0.090968718 0.52559836
[30,35)  0.0034040485 0.0025475576 0.008310784 0.065447652 0.34975552
[35,40)  0.0035764314 0.0025109233 0.007379301 0.051192518 0.25155153
[40,45)  0.0036871730 0.0025128297 0.007036797 0.045837821 0.20914486
[45,50)  0.0040915085 0.0027227246 0.007306960 0.045923901 0.19896364
[50,55)  0.0043095498 0.0028109059 0.007162739 0.043173177 0.18013096
[55,60)  0.0036274422 0.0023972811 0.005795537 0.032911798 0.13219103
[60,65)  0.0021681315 0.0015802011 0.003795463 0.020101314 0.07483048
[65,70)  0.0009439189 0.0008293203 0.002205160 0.011364312 0.03747182
[70,75)  0.0003624376 0.0003764863 0.001254462 0.007157981 0.02251799
[75,Inf) 0.0003319858 0.0002569713 0.001046544 0.008773102 0.03605186
            [25,30)    [30,35)    [35,40)    [40,45)    [45,50)    [50,55)
[0,5)    0.08541531 0.08574137 0.08075343 0.07799741 0.07073843 0.05537374
[5,10)   0.10848344 0.11413945 0.11188038 0.11211256 0.10713263 0.09219048
[10,15)  0.14857096 0.16021581 0.16064896 0.16063819 0.15454987 0.13990863
[15,20)  0.23663570 0.23943721 0.24840671 0.24513567 0.22592498 0.20144928
[20,25)  0.50944238 0.41013017 0.42767962 0.42449595 0.37401441 0.31270634
[25,30)  0.90384900 0.65049771 0.62731445 0.62608717 0.54060631 0.43003526
[30,35)  0.72174430 0.76563365 0.71009033 0.70271454 0.59921273 0.46830980
[35,40)  0.50782207 0.60208714 0.67436300 0.68683499 0.58261471 0.44900894
[40,45)  0.40989766 0.48090633 0.56225822 0.65761289 0.58845796 0.44639398
[45,50)  0.37587745 0.44146986 0.50278802 0.58409870 0.62661604 0.50307875
[50,55)  0.32449770 0.37879774 0.43321902 0.47880584 0.50758442 0.51659659
[55,60)  0.22805693 0.25849245 0.30116161 0.33026505 0.32569678 0.33340526
[60,65)  0.12248341 0.13432304 0.15716717 0.17719763 0.16863239 0.15922927
[65,70)  0.05620608 0.05978026 0.07028509 0.08217683 0.07899226 0.06974852
[70,75)  0.03082137 0.03121590 0.03722263 0.04599962 0.04596189 0.03898958
[75,Inf) 0.05335541 0.05059575 0.05830441 0.08151596 0.09980679 0.09105309
            [55,60)    [60,65)     [65,70)     [70,75)    [75,Inf)
[0,5)    0.03055127 0.01069179 0.003563600 0.002505352 0.003870799
[5,10)   0.05933059 0.02386746 0.007791835 0.004004593 0.003652821
[10,15)  0.10095239 0.04800492 0.017638952 0.008149083 0.004192452
[15,20)  0.15203986 0.08183562 0.035953414 0.018469849 0.006827851
[20,25)  0.22681893 0.12385324 0.059426914 0.036587025 0.013436401
[25,30)  0.28831062 0.14570531 0.069279246 0.048963671 0.021379300
[30,35)  0.30265656 0.13882810 0.059851566 0.042479168 0.022066690
[35,40)  0.29509387 0.13132425 0.049731355 0.030766282 0.016394444
[40,45)  0.29780623 0.13938490 0.050355057 0.026140792 0.011839894
[45,50)  0.33204169 0.16317160 0.061726550 0.028894226 0.010248698
[50,55)  0.35996337 0.17841733 0.071107803 0.032883356 0.009457324
[55,60)  0.30333468 0.15947907 0.064521417 0.030085638 0.007621199
[60,65)  0.14935678 0.10368239 0.045752391 0.021392984 0.004924129
[65,70)  0.06093550 0.04521728 0.027697711 0.014236741 0.003009743
[70,75)  0.03124344 0.02227795 0.014638277 0.010420579 0.002363574
[75,Inf) 0.06113815 0.03526483 0.022060067 0.016386764 0.005709931

$school
                [0,5)      [5,10)     [10,15)     [15,20)      [20,25)
[0,5)    1.2358696998 0.247486320 0.032193829 0.020850528 0.0366915825
[5,10)   0.3179310607 4.498408402 0.342129199 0.032837759 0.0394399848
[10,15)  0.0755278792 0.412354577 6.602110276 0.440572952 0.0700414646
[15,20)  0.0572976718 0.080511140 0.558165344 5.414594331 0.3463262272
[20,25)  0.0765402818 0.102972389 0.166516239 0.503161773 0.8412040183
[25,30)  0.1034619161 0.130279917 0.126084138 0.150040873 0.2169940136
[30,35)  0.1162951590 0.173128814 0.165164646 0.114578965 0.0640173828
[35,40)  0.0988653023 0.175737694 0.199610818 0.140615391 0.0434605441
[40,45)  0.0726482006 0.147718028 0.187944765 0.166042420 0.0521564679
[45,50)  0.0540968969 0.124849158 0.160628204 0.157667068 0.0650820375
[50,55)  0.0390590271 0.101063040 0.130938665 0.118473281 0.0567291283
[55,60)  0.0231998151 0.063427822 0.086200976 0.069251563 0.0305800143
[60,65)  0.0099569993 0.026856808 0.037893943 0.029584381 0.0129131164
[65,70)  0.0031529497 0.008300227 0.011848754 0.009096099 0.0052756383
[70,75)  0.0009529908 0.002383149 0.003245067 0.002172102 0.0012399827
[75,Inf) 0.0015920412 0.002589751 0.002491511 0.001053170 0.0003551419
              [25,30)      [30,35)     [35,40)     [40,45)     [45,50)
[0,5)    0.0646516722 0.0640423739 0.047315981 0.060371758 0.101896890
[5,10)   0.0710216181 0.1153472661 0.109600111 0.121213994 0.207901620
[10,15)  0.0552525717 0.1046319274 0.142556457 0.150273858 0.206484573
[15,20)  0.0718415846 0.0743912189 0.123642828 0.150202705 0.162172367
[20,25)  0.1486215645 0.0465312628 0.053981076 0.074231183 0.076299523
[25,30)  0.2961808482 0.0818562792 0.040460944 0.046800565 0.050215344
[30,35)  0.1066361894 0.1622009398 0.061649782 0.047742449 0.048127062
[35,40)  0.0314694997 0.0630760101 0.071094589 0.048578686 0.046367621
[40,45)  0.0203285215 0.0230012722 0.034078503 0.039934367 0.045420631
[45,50)  0.0252626911 0.0164908729 0.017485770 0.024043071 0.041321567
[50,55)  0.0298387752 0.0186131356 0.013190196 0.013794849 0.023680207
[55,60)  0.0203295413 0.0167506823 0.011430455 0.008564327 0.009996071
[60,65)  0.0096119492 0.0109345007 0.009726346 0.007078658 0.005452022
[65,70)  0.0046486152 0.0064222712 0.008317668 0.007534387 0.005131395
[70,75)  0.0010920323 0.0016911875 0.003033472 0.003759291 0.002741524
[75,Inf) 0.0002347126 0.0003985183 0.001108305 0.002566262 0.002936996
             [50,55)      [55,60)      [60,65)      [65,70)      [70,75)
[0,5)    0.111540917 0.0689229326 0.0303233734 0.0100116875 0.0022718430
[5,10)   0.268349879 0.1528186883 0.0492270517 0.0168105904 0.0070864370
[10,15)  0.276743106 0.1719292437 0.0467756630 0.0126054097 0.0071180504
[15,20)  0.180090388 0.1225171357 0.0344122268 0.0067115699 0.0029223018
[20,25)  0.066381028 0.0443558711 0.0164280754 0.0039186747 0.0010282893
[25,30)  0.038707054 0.0239966116 0.0119549148 0.0038528858 0.0006233083
[30,35)  0.035452559 0.0210155654 0.0131419956 0.0059015177 0.0007926006
[35,40)  0.032048855 0.0172626997 0.0131270183 0.0098082738 0.0016822804
[40,45)  0.033636435 0.0153681201 0.0114284413 0.0142364478 0.0043316022
[45,50)  0.044416123 0.0195019024 0.0104557120 0.0162572310 0.0097638800
[50,55)  0.051143630 0.0333379479 0.0125009921 0.0140967349 0.0133427018
[55,60)  0.021116413 0.0402120602 0.0193401479 0.0113680788 0.0096515725
[60,65)  0.006670482 0.0138593883 0.0248064365 0.0111717648 0.0049073717
[65,70)  0.003674495 0.0044116127 0.0080533915 0.0094131841 0.0028034789
[70,75)  0.001519927 0.0011136884 0.0013891506 0.0019236740 0.0016463565
[75,Inf) 0.001533895 0.0006295928 0.0004720087 0.0005558011 0.0005999646
             [75,Inf)
[0,5)    0.0002418317
[5,10)   0.0017264725
[10,15)  0.0053662254
[15,20)  0.0076887406
[20,25)  0.0064600058
[25,30)  0.0029941645
[30,35)  0.0007293494
[35,40)  0.0001878398
[40,45)  0.0002459017
[45,50)  0.0009214829
[50,55)  0.0029394245
[55,60)  0.0046223541
[60,65)  0.0033614790
[65,70)  0.0014681503
[70,75)  0.0005724311
[75,Inf) 0.0004628103

$other
              [0,5)     [5,10)    [10,15)    [15,20)    [20,25)    [25,30)
[0,5)    0.81478145 0.45256395 0.10917264 0.07894378 0.12307902 0.16430265
[5,10)   0.41931103 1.82170694 0.64279606 0.14369974 0.12559994 0.17320691
[10,15)  0.15456026 0.70094600 2.94289487 0.77429570 0.20212198 0.19275555
[15,20)  0.14466070 0.20363249 0.97503511 3.28269950 0.79289273 0.29139412
[20,25)  0.24961582 0.17730485 0.27027619 1.18360196 2.65007558 0.78197436
[25,30)  0.41873117 0.27419642 0.20502680 0.38098229 1.17396320 1.61008983
[30,35)  0.46883939 0.40232329 0.24325593 0.24579509 0.46123947 0.85475513
[35,40)  0.34794320 0.40489139 0.30385643 0.23854219 0.29109001 0.42489448
[40,45)  0.23220288 0.28696080 0.29001756 0.26229359 0.26163860 0.29686619
[45,50)  0.19255317 0.19112428 0.21122870 0.24643244 0.27369098 0.28367683
[50,55)  0.20140850 0.15035891 0.13726777 0.17746039 0.25023977 0.29465932
[55,60)  0.22462026 0.14589723 0.09918483 0.10855714 0.17585721 0.26386805
[60,65)  0.21180314 0.14335045 0.08667867 0.06948485 0.09749767 0.17018673
[65,70)  0.15034532 0.11690769 0.07757210 0.05286184 0.05150181 0.07920448
[70,75)  0.08370649 0.07438835 0.06007261 0.04410846 0.03429442 0.03781846
[75,Inf) 0.06042292 0.05775048 0.06074408 0.06618968 0.06050986 0.05455752
            [30,35)    [35,40)   [40,45)    [45,50)    [50,55)    [55,60)
[0,5)    0.17472242 0.15872293 0.1144962 0.08582814 0.08322425 0.08246880
[5,10)   0.20972186 0.22996422 0.1889557 0.12351565 0.09375080 0.09797056
[10,15)  0.22910965 0.26326858 0.2694367 0.20954175 0.13945239 0.11489136
[15,20)  0.27860098 0.28845443 0.3099640 0.31712781 0.25531656 0.17729369
[20,25)  0.44629916 0.39514898 0.3559339 0.40098547 0.44664444 0.35197931
[25,30)  0.77370274 0.58409393 0.4342284 0.39971265 0.53025088 0.56652203
[30,35)  0.93591372 0.72538589 0.5327966 0.37570996 0.42985935 0.57362723
[35,40)  0.58693853 0.66873042 0.5962530 0.40732499 0.34338549 0.43347243
[40,45)  0.35349900 0.46132728 0.5543347 0.48119970 0.35869798 0.35219448
[45,50)  0.28377089 0.31481319 0.4103487 0.51782967 0.46612993 0.40021291
[50,55)  0.28939396 0.25968271 0.2797822 0.40041194 0.53679210 0.53305087
[55,60)  0.31070509 0.26950049 0.2158998 0.25804370 0.43280103 0.59454555
[60,65)  0.26311901 0.27953644 0.1977481 0.16614160 0.25470577 0.45295014
[65,70)  0.14670597 0.21311292 0.1799328 0.12278189 0.13495249 0.23580024
[70,75)  0.06489085 0.11622944 0.1380154 0.10696603 0.09079191 0.11962904
[75,Inf) 0.06363760 0.09369326 0.1325008 0.15054708 0.15324037 0.16243975
            [60,65)    [65,70)    [70,75)   [75,Inf)
[0,5)    0.06765329 0.05644393 0.06925855 0.07257282
[5,10)   0.10434401 0.09114718 0.08051756 0.07631303
[10,15)  0.12640516 0.13213018 0.10524390 0.05971995
[15,20)  0.15335093 0.16876950 0.15600388 0.05931031
[20,25)  0.23890845 0.21023350 0.21956983 0.08662189
[25,30)  0.38792043 0.26404222 0.25797758 0.12422600
[30,35)  0.49244221 0.31550591 0.26187355 0.13881252
[35,40)  0.45962454 0.33766026 0.26329742 0.13444629
[40,45)  0.37395422 0.32047334 0.27502788 0.13890602
[45,50)  0.36110078 0.30731513 0.28869915 0.16681305
[50,55)  0.45070226 0.34145651 0.29952552 0.20156110
[55,60)  0.58763639 0.44385525 0.33430965 0.21850540
[60,65)  0.58360793 0.54767780 0.39987158 0.20850116
[65,70)  0.38002664 0.50718751 0.46538519 0.19239406
[70,75)  0.19124072 0.29183395 0.43765109 0.20005401
[75,Inf) 0.18816965 0.23910618 0.35112878 0.39091971

$all
             [0,5)    [5,10)    [10,15)    [15,20)    [20,25)    [25,30)
[0,5)    2.6194968 1.0893795  0.3233530 0.22896181 0.38379538 0.69194745
[5,10)   1.1918875 7.0725175  1.4471920 0.37150683 0.34521899 0.55696339
[10,15)  0.5064876 1.6037301 10.3962076 1.69852692 0.53371388 0.51403721
[15,20)  0.4320243 0.5146815  1.9974643 9.60582092 1.75597064 0.74091401
[20,25)  0.6605047 0.4607658  0.6376974 2.20370267 4.78998765 1.79518660
[25,30)  1.0927157 0.6756063  0.4830084 0.76424247 2.24738399 3.48971751
[30,35)  1.2512859 1.0560966  0.6392382 0.52842969 0.98722983 1.98484955
[35,40)  0.8976938 1.1561324  0.9293785 0.59611030 0.66270670 1.06252931
[40,45)  0.5379418 0.8233874  1.0083140 0.80826333 0.65100606 0.79346840
[45,50)  0.4072469 0.5140882  0.7388462 0.89384074 0.81523279 0.79994723
[50,55)  0.4111453 0.3797018  0.4440904 0.63532275 0.85200082 0.89142637
[55,60)  0.4399385 0.3378610  0.2854941 0.33605136 0.56505215 0.80561779
[60,65)  0.3819526 0.3077217  0.2105421 0.17915228 0.26995535 0.46227710
[65,70)  0.2501437 0.2343943  0.1712034 0.11911803 0.13040945 0.19394412
[70,75)  0.1379713 0.1412109  0.1254746 0.09572899 0.08524679 0.09388003
[75,Inf) 0.1126203 0.1223715  0.1268703 0.13327569 0.14933130 0.15771897
           [30,35)   [35,40)   [40,45)   [45,50)   [50,55)   [55,60)   [60,65)
[0,5)    0.8536773 0.6460735 0.4222595 0.3736569 0.3864045 0.3616037 0.2816777
[5,10)   0.9224997 1.1071588 0.8365958 0.6321491 0.5939343 0.4811791 0.3869020
[10,15)  0.7307321 1.1097509 1.2401475 0.9579246 0.7437699 0.5293386 0.3859341
[15,20)  0.7121243 0.9032815 1.1983916 1.2370304 0.9452866 0.6094109 0.3875233
[20,25)  1.0446636 1.0029908 1.0881571 1.2650332 1.2473179 0.8796617 0.5139167
[25,30)  1.8463824 1.3931739 1.2331953 1.2007381 1.3387141 1.2307878 0.7685713
[30,35)  2.4624541 1.7893233 1.4138314 1.1399886 1.1132949 1.1810805 0.9435108
[35,40)  1.5207999 1.8856180 1.5737579 1.1584049 0.9329123 0.9049823 0.8501797
[40,45)  0.9525148 1.2862255 1.6314584 1.3428045 0.9629746 0.7728176 0.6737228
[45,50)  0.8107096 0.9299928 1.2234180 1.5647073 1.2703053 0.8926446 0.6490256
[50,55)  0.8014576 0.7786774 0.8634086 1.1366551 1.5441805 1.2319048 0.7992597
[55,60)  0.8021330 0.6926241 0.6224678 0.6781457 1.0058683 1.4473929 1.0971037
[60,65)  0.6329088 0.6225748 0.4713188 0.3946175 0.4958286 0.8369486 1.2029913
[65,70)  0.3207948 0.4505583 0.3923286 0.2680031 0.2484378 0.3654160 0.6254597
[70,75)  0.1352022 0.2326635 0.2973072 0.2372403 0.1734808 0.1839119 0.2713977
[75,Inf) 0.1626995 0.2076237 0.3012825 0.3758589 0.3706038 0.3255729 0.3058989
           [65,70)   [70,75)  [75,Inf)
[0,5)    0.1931504 0.1723937 0.2070321
[5,10)   0.2939867 0.2198456 0.2440371
[10,15)  0.3436658 0.2761259 0.2372748
[15,20)  0.3387182 0.3181326 0.2328029
[20,25)  0.3697879 0.3604127 0.2585333
[25,30)  0.4544454 0.3934787 0.2905082
[30,35)  0.5732568 0.4131476 0.2888511
[35,40)  0.6527624 0.4703171 0.2770188
[40,45)  0.6019504 0.5367344 0.3125663
[45,50)  0.5298609 0.5336493 0.4071881
[50,55)  0.5438098 0.4875705 0.4845223
[55,60)  0.6777572 0.4881291 0.4786380
[60,65)  0.9006746 0.5689396 0.3963394
[65,70)  0.9381085 0.7218211 0.3292546
[70,75)  0.4681275 0.7530014 0.3503535
[75,Inf) 0.3574148 0.5629846 0.7994623

plot new contact matrix

plot_matrix(synthetic_fairfield$home)

plot POLYMOD

plot_matrix(synthetic_polymod$home)

Challenges in Design

synthetic_fairfield <- extrapolate_polymod(
 population = fairfield,
 age_breaks = c(seq(0, 75, by = 5), Inf)
)

• Model fitting is fixed and rigid
  • Although there are arguments for age_break
• Design decision to provide access to fitting the model
• Fixed and rigid means you cannot change the

Challenges in Design

• Hard coding variables
  • age and population columns are currently required to be named lower.age.limit and population
  • At least 2 potential ways to resolve this

fairfield

# A tibble: 18 × 4
   lga           lower.age.limit  year population
   <chr>                   <dbl> <dbl>      <dbl>
 1 Fairfield (C)               0  2020      12261
 2 Fairfield (C)               5  2020      13093
 3 Fairfield (C)              10  2020      13602
 4 Fairfield (C)              15  2020      14323
 5 Fairfield (C)              20  2020      15932
 6 Fairfield (C)              25  2020      16190
 7 Fairfield (C)              30  2020      14134
 8 Fairfield (C)              35  2020      13034
 9 Fairfield (C)              40  2020      12217
10 Fairfield (C)              45  2020      13449
11 Fairfield (C)              50  2020      13419
12 Fairfield (C)              55  2020      13652
13 Fairfield (C)              60  2020      12907
14 Fairfield (C)              65  2020      10541
15 Fairfield (C)              70  2020       8227
16 Fairfield (C)              75  2020       5598
17 Fairfield (C)              80  2020       4006
18 Fairfield (C)              85  2020       4240

Challenges in Design

1. Provide arguments for specifying the age and population information (age_col and population_col)

extrapolate_polymod(
 population = fairfield,
 age_col = lower.age.limit,
 population_col = population
)

Challenges in Design

2. Some kind of class based approach, where we encode or validate the age and population columns.

fairfield_pop <- conmat_population(
  data = fairfield,
  age = lower.age.limit,
  population = population
)
extrapolate_polymod(
 population = fairfield_pop
)

Challenges in Design

• Rigid model design means easy to use, at cost of flexibility
• Specified variable names is fragile, but saves typing
• Extra functions could always be added to be more flexible variants

Future Directions

• Interface to changing GAM model terms or specify model
• Better automatic plotting
• Validating against Prem and other methods
• Allowing more flexible uses of other data sources

Thanks

• Nick Golding
• Aarathy Babu
• Michael Lydeamore

Learning more

conmat package

njtierney.github.io/talk-nzrse-2022

nj_tierney

njtierney

nicholas.tierney@gmail.com

End.