EDA_Vaccinations

I would like to explore the bottom postcodes for low immun % coverage and the top postcodes for immun % coverage

This task was extremely hard, as I could only identify bottom 20 postcodes and top 20 postcodes, but due to duplicates I could then only explore the bottom class of immunisation % coverage (<70% and >95%) and the top class.

#load libraries
library(readr)
library(ggplot2)
library(dplyr)

## 
## Attaching package: 'dplyr'

## The following objects are masked from 'package:stats':
## 
##     filter, lag

## The following objects are masked from 'package:base':
## 
##     intersect, setdiff, setequal, union

library(Amelia)

## Loading required package: Rcpp

## ## 
## ## Amelia II: Multiple Imputation
## ## (Version 1.7.5, built: 2018-05-07)
## ## Copyright (C) 2005-2018 James Honaker, Gary King and Matthew Blackwell
## ## Refer to http://gking.harvard.edu/amelia/ for more information
## ##

library(reshape2)

Immport and clean data

I used the merge file containing, PHN, postcode, immun%coverage, SEIFA data and electoral data. (IE, there were no demographics or taxable income data in this file)

#import newly merged dataset
immunisations_merged <- read.csv("cleaned_data/immunization_with_everything.csv")
str(immunisations_merged)

## 'data.frame':    191163 obs. of  46 variables:
##  $ X.1                                        : int  1 2 3 4 5 6 7 8 9 10 ...
##  $ postcode                                   : int  800 800 800 800 800 800 800 800 800 800 ...
##  $ X                                          : int  163657 132679 132628 132622 132677 132621 163658 132678 163662 163663 ...
##  $ state                                      : Factor w/ 12 levels "ACT","NSW","NSW/ACT",..: 6 6 6 6 6 6 6 6 6 6 ...
##  $ year                                       : int  2016 2011 2014 2012 2011 2012 2016 2011 2016 2016 ...
##  $ age                                        : int  5 1 1 1 1 1 2 1 1 1 ...
##  $ pc_immun                                   : Factor w/ 9 levels "<70.0","70.0-74.9",..: 6 6 7 7 6 7 3 6 6 6 ...
##  $ caution                                    : int  1 1 1 1 1 1 1 1 1 1 ...
##  $ pc_immun_class                             : int  6 6 7 7 6 7 3 6 6 6 ...
##  $ PHN_code                                   : Factor w/ 31 levels "PHN101","PHN102",..: 30 30 30 30 30 30 30 30 30 30 ...
##  $ PHN_number                                 : int  701 701 701 701 701 701 701 701 701 701 ...
##  $ Index.type                                 : Factor w/ 4 levels "Index of Economic Resources",..: 4 4 2 3 1 1 3 3 3 2 ...
##  $ Time                                       : int  2016 2011 2011 2011 2011 2011 2016 2011 2016 2016 ...
##  $ Maximum.score.for.SA1s.in.area             : int  NA 1127 1167 1144 986 986 1167 1144 1167 1163 ...
##  $ Minimum.score.for.SA1s.in.area             : int  842 1032 1066 1051 927 927 914 1051 914 1023 ...
##  $ Rank.within.Australia                      : int  2243 2053 2089 2136 501 501 2398 2136 2398 2287 ...
##  $ Rank.within.Australia...Decile             : int  NA 9 9 9 3 3 10 9 10 9 ...
##  $ Rank.within.Australia...Percentile         : int  86 83 85 87 21 21 92 87 92 87 ...
##  $ Rank.within.State.or.Territory             : int  28 25 28 26 15 15 33 26 33 33 ...
##  $ Rank.within.State.or.Territory...Decile    : int  8 9 10 9 6 6 10 9 10 10 ...
##  $ Rank.within.State.or.Territory...Percentile: int  NA NA NA NA NA NA 92 NA 92 92 ...
##  $ Score                                      : int  1066 1060 1077 1072 952 952 1096 1072 1096 1089 ...
##  $ Usual.resident.population                  : int  6464 0 0 4564 0 0 6464 4564 6464 6464 ...
##  $ Electoral.division                         : Factor w/ 150 levels "Adelaide","Aston",..: 136 136 136 136 136 136 136 136 136 136 ...
##  $ Per.cent.postcode.in.electorate            : num  100 100 100 100 100 100 100 100 100 100 ...
##  $ DivisionID2016                             : int  307 307 307 307 307 307 307 307 307 307 ...
##  $ StateAB2016                                : Factor w/ 8 levels "ACT","NSW","NT",..: 3 3 3 3 3 3 3 3 3 3 ...
##  $ CandidateID2016                            : int  28737 28737 28737 28737 28737 28737 28737 28737 28737 28737 ...
##  $ GivenNm2016                                : Factor w/ 97 levels "Adam","Alan",..: 56 56 56 56 56 56 56 56 56 56 ...
##  $ Surname2016                                : Factor w/ 135 levels "ABBOTT","ALBANESE",..: 51 51 51 51 51 51 51 51 51 51 ...
##  $ PartyNm2016                                : Factor w/ 8 levels "Australian Labor Party",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ PartyAb2016                                : Factor w/ 8 levels "ALP","GRN","IND",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ DivisionID2010                             : int  307 307 307 307 307 307 307 307 307 307 ...
##  $ StateAb2010                                : Factor w/ 8 levels "ACT","NSW","NT",..: 3 3 3 3 3 3 3 3 3 3 ...
##  $ CandidateID2010                            : int  21477 21477 21477 21477 21477 21477 21477 21477 21477 21477 ...
##  $ GivenNm2010                                : Factor w/ 108 levels "Adam","Alan",..: 74 74 74 74 74 74 74 74 74 74 ...
##  $ Surname2010                                : Factor w/ 137 levels "ABBOTT","ADAMS",..: 52 52 52 52 52 52 52 52 52 52 ...
##  $ PartyNm2010                                : Factor w/ 7 levels "Australian Labor Party",..: 2 2 2 2 2 2 2 2 2 2 ...
##  $ PartyAb2010                                : Factor w/ 7 levels "ALP","CLP","GRN",..: 2 2 2 2 2 2 2 2 2 2 ...
##  $ DivisionID2013                             : int  307 307 307 307 307 307 307 307 307 307 ...
##  $ StateAb2013                                : Factor w/ 8 levels "ACT","NSW","NT",..: 3 3 3 3 3 3 3 3 3 3 ...
##  $ CandidateID2013                            : int  23502 23502 23502 23502 23502 23502 23502 23502 23502 23502 ...
##  $ GivenNm2013                                : Factor w/ 106 levels "Adam","Alan",..: 76 76 76 76 76 76 76 76 76 76 ...
##  $ Surname2013                                : Factor w/ 140 levels "ABBOTT","ALBANESE",..: 48 48 48 48 48 48 48 48 48 48 ...
##  $ PartyNm2013                                : Factor w/ 9 levels "Australian Labor Party",..: 2 2 2 2 2 2 2 2 2 2 ...
##  $ PartyAb2013                                : Factor w/ 9 levels "ALP","CLP","GRN",..: 2 2 2 2 2 2 2 2 2 2 ...

#check to see how it really looks - it's a mix of long and wide data
summary(immunisations_merged)

##       X.1            postcode          X              state      
##  Min.   :     1   Min.   : 800   Min.   :     1   VIC    :50376  
##  1st Qu.: 47792   1st Qu.:2774   1st Qu.: 39310   NSW    :49864  
##  Median : 95582   Median :3788   Median : 78666   QLD    :33463  
##  Mean   : 95582   Mean   :4001   Mean   : 80160   SA     :22900  
##  3rd Qu.:143372   3rd Qu.:5070   3rd Qu.:121512   WA     :21948  
##  Max.   :191163   Max.   :7470   Max.   :164021   TAS    : 7856  
##                                                   (Other): 4756  
##       year           age              pc_immun        caution      
##  Min.   :2011   Min.   :1.000   NP        :58819   Min.   :0.0000  
##  1st Qu.:2012   1st Qu.:1.000   92.5-94.9 :35652   1st Qu.:0.0000  
##  Median :2014   Median :2.000   90.0-92.4 :33580   Median :0.0000  
##  Mean   :2014   Mean   :2.671   85.0-89.9 :27468   Mean   :0.2413  
##  3rd Qu.:2015   3rd Qu.:5.000   95.0-100.0:27160   3rd Qu.:0.0000  
##  Max.   :2016   Max.   :5.000   80.0-84.9 : 5936   Max.   :1.0000  
##                                 (Other)   : 2548                   
##  pc_immun_class     PHN_code        PHN_number   
##  Min.   :0.000   PHN402 : 13028   Min.   :101.0  
##  1st Qu.:0.000   PHN503 : 11884   1st Qu.:110.0  
##  Median :6.000   PHN205 : 10888   Median :206.0  
##  Mean   :4.372   PHN202 : 10092   Mean   :280.8  
##  3rd Qu.:7.000   PHN401 :  9872   3rd Qu.:401.0  
##  Max.   :8.000   PHN206 :  9760   Max.   :801.0  
##                  (Other):125639                  
##                                                        Index.type   
##  Index of Economic Resources                                :47790  
##  Index of Education and Occupation                          :47790  
##  Index of Relative Socio-economic Advantage and Disadvantage:47790  
##  Index of Relative Socio-economic Disadvantage              :47793  
##                                                                     
##                                                                     
##                                                                     
##       Time      Maximum.score.for.SA1s.in.area
##  Min.   :2011   Min.   : 506                  
##  1st Qu.:2011   1st Qu.:1030                  
##  Median :2011   Median :1080                  
##  Mean   :2012   Mean   :1077                  
##  3rd Qu.:2011   3rd Qu.:1129                  
##  Max.   :2016   Max.   :1375                  
##  NA's   :3      NA's   :4826                  
##  Minimum.score.for.SA1s.in.area Rank.within.Australia
##  Min.   : 121.0                 Min.   :   1         
##  1st Qu.: 818.0                 1st Qu.: 651         
##  Median : 897.0                 Median :1288         
##  Mean   : 882.9                 Mean   :1276         
##  3rd Qu.: 969.0                 3rd Qu.:1906         
##  Max.   :1226.0                 Max.   :2630         
##  NA's   :4978                   NA's   :4812         
##  Rank.within.Australia...Decile Rank.within.Australia...Percentile
##  Min.   : 1.000                 Min.   :  1.00                    
##  1st Qu.: 3.000                 1st Qu.: 27.00                    
##  Median : 6.000                 Median : 52.00                    
##  Mean   : 5.592                 Mean   : 51.45                    
##  3rd Qu.: 8.000                 3rd Qu.: 77.00                    
##  Max.   :10.000                 Max.   :100.00                    
##  NA's   :4777                   NA's   :4782                      
##  Rank.within.State.or.Territory Rank.within.State.or.Territory...Decile
##  Min.   :  1.0                  Min.   : 1.000                         
##  1st Qu.:102.0                  1st Qu.: 3.000                         
##  Median :230.0                  Median : 6.000                         
##  Mean   :250.7                  Mean   : 5.615                         
##  3rd Qu.:370.0                  3rd Qu.: 8.000                         
##  Max.   :694.0                  Max.   :10.000                         
##  NA's   :4797                   NA's   :4838                           
##  Rank.within.State.or.Territory...Percentile     Score       
##  Min.   :  1.00                              Min.   : 506.0  
##  1st Qu.: 27.00                              1st Qu.: 947.0  
##  Median : 52.00                              Median : 993.0  
##  Mean   : 51.65                              Mean   : 993.9  
##  3rd Qu.: 77.00                              3rd Qu.:1043.0  
##  Max.   :100.00                              Max.   :1234.0  
##  NA's   :8123                                NA's   :4718    
##  Usual.resident.population Electoral.division
##  Min.   :     0            O'Connor:  5108   
##  1st Qu.:     0            Maranoa :  4959   
##  Median :     0            Grey    :  4852   
##  Mean   :  4140            Mallee  :  3892   
##  3rd Qu.:  2112            Wannon  :  3692   
##  Max.   :106745            Barker  :  3656   
##  NA's   :4829              (Other) :165004   
##  Per.cent.postcode.in.electorate DivisionID2016   StateAB2016   
##  Min.   :  0.1                   Min.   :101.0   VIC    :50316  
##  1st Qu.: 91.9                   1st Qu.:158.0   NSW    :44164  
##  Median :100.0                   Median :190.0   QLD    :32499  
##  Mean   : 84.8                   Mean   :190.7   SA     :22972  
##  3rd Qu.:100.0                   3rd Qu.:226.0   WA     :18596  
##  Max.   :100.0                   Max.   :310.0   (Other):10940  
##                                  NA's   :11676   NA's   :11676  
##  CandidateID2016  GivenNm2016          Surname2016    
##  Min.   :20881   Andrew :  9896   WILSON     :  6580  
##  1st Qu.:28317   Tony   :  7744   MITCHELL   :  6576  
##  Median :28777   David  :  6919   LITTLEPROUD:  4959  
##  Mean   :28620   Rick   :  5108   RAMSEY     :  4852  
##  3rd Qu.:28997   Rowan  :  4852   BROAD      :  3892  
##  Max.   :29576   (Other):144968   (Other)    :152628  
##  NA's   :11676   NA's   : 11676   NA's       : 11676  
##                  PartyNm2016     PartyAb2016    DivisionID2010 
##  Australian Labor Party:65980   ALP    :65980   Min.   :101.0  
##  Liberal               :61084   LP     :61084   1st Qu.:156.0  
##  Liberal National Party:21935   LNP    :21935   Median :189.0  
##  The Nationals         :19584   NP     :19584   Mean   :189.5  
##  Independent           : 3864   IND    : 3864   3rd Qu.:226.0  
##  (Other)               : 7040   (Other): 7040   Max.   :310.0  
##  NA's                  :11676   NA's   :11676   NA's   :8908   
##   StateAb2010    CandidateID2010  GivenNm2010      Surname2010    
##  VIC    :50316   Min.   :13270   Tony   : 11012   CROOK  :  5108  
##  NSW    :45740   1st Qu.:20876   John   :  6780   SCOTT  :  4959  
##  QLD    :32499   Median :21088   Bruce  :  5535   RAMSEY :  4852  
##  SA     :22972   Mean   :20988   Darren :  4940   FORREST:  3892  
##  WA     :18596   3rd Qu.:21416   Andrew :  4908   TEHAN  :  3692  
##  (Other):12132   Max.   :21912   (Other):149080   (Other):159752  
##  NA's   : 8908   NA's   :8908    NA's   :  8908   NA's   :  8908  
##                                PartyNm2010     PartyAb2010   
##  Australian Labor Party              :73548   ALP    :73548  
##  Liberal                             :61948   LP     :61948  
##  Liberal National Party of Queensland:21135   LNQ    :21135  
##  The Nationals                       :17868   NP     :17868  
##  Independent                         : 5912   IND    : 5912  
##  (Other)                             : 1844   (Other): 1844  
##  NA's                                : 8908   NA's   : 8908  
##  DivisionID2013   StateAb2013    CandidateID2013  GivenNm2013    
##  Min.   :101.0   VIC    :50316   Min.   :20881   Andrew : 10772  
##  1st Qu.:156.0   NSW    :45740   1st Qu.:23312   Tony   :  7744  
##  Median :189.0   QLD    :32499   Median :23656   Bruce  :  5535  
##  Mean   :189.5   SA     :22972   Mean   :23675   Rick   :  5108  
##  3rd Qu.:226.0   WA     :18596   3rd Qu.:23903   Rowan  :  4852  
##  Max.   :310.0   (Other):12132   Max.   :24739   (Other):148244  
##  NA's   :8908    NA's   : 8908   NA's   :8908    NA's   :  8908  
##   Surname2013                     PartyNm2013     PartyAb2013   
##  SCOTT  :  5671   Liberal               :80548   LP     :80548  
##  WILSON :  5108   Australian Labor Party:53352   ALP    :53352  
##  RAMSEY :  4852   Liberal National Party:22667   LNP    :22667  
##  BROAD  :  3892   The Nationals         :16568   NP     :16568  
##  TEHAN  :  3692   Independent           : 3864   IND    : 3864  
##  (Other):159040   (Other)               : 5256   (Other): 5256  
##  NA's   :  8908   NA's                  : 8908   NA's   : 8908

1. Reduced columns names

# Clean data using function 

# Parameters
# * imm_merged - the source data frame we want to clean

# Returns - a data frame

data <- immunisations_merged %>%
    select('state', 'postcode', 'year', 'age',  'pc_immun', 'caution', 'pc_immun_class','PHN_code', 'PHN_number','Usual.resident.population','Index.type','Score', 'Electoral.division', 'PartyNm2016', 'PartyNm2010', 'PartyNm2013')

data$age <- as.factor(data$age)

str(data)

## 'data.frame':    191163 obs. of  16 variables:
##  $ state                    : Factor w/ 12 levels "ACT","NSW","NSW/ACT",..: 6 6 6 6 6 6 6 6 6 6 ...
##  $ postcode                 : int  800 800 800 800 800 800 800 800 800 800 ...
##  $ year                     : int  2016 2011 2014 2012 2011 2012 2016 2011 2016 2016 ...
##  $ age                      : Factor w/ 3 levels "1","2","5": 3 1 1 1 1 1 2 1 1 1 ...
##  $ pc_immun                 : Factor w/ 9 levels "<70.0","70.0-74.9",..: 6 6 7 7 6 7 3 6 6 6 ...
##  $ caution                  : int  1 1 1 1 1 1 1 1 1 1 ...
##  $ pc_immun_class           : int  6 6 7 7 6 7 3 6 6 6 ...
##  $ PHN_code                 : Factor w/ 31 levels "PHN101","PHN102",..: 30 30 30 30 30 30 30 30 30 30 ...
##  $ PHN_number               : int  701 701 701 701 701 701 701 701 701 701 ...
##  $ Usual.resident.population: int  6464 0 0 4564 0 0 6464 4564 6464 6464 ...
##  $ Index.type               : Factor w/ 4 levels "Index of Economic Resources",..: 4 4 2 3 1 1 3 3 3 2 ...
##  $ Score                    : int  1066 1060 1077 1072 952 952 1096 1072 1096 1089 ...
##  $ Electoral.division       : Factor w/ 150 levels "Adelaide","Aston",..: 136 136 136 136 136 136 136 136 136 136 ...
##  $ PartyNm2016              : Factor w/ 8 levels "Australian Labor Party",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ PartyNm2010              : Factor w/ 7 levels "Australian Labor Party",..: 2 2 2 2 2 2 2 2 2 2 ...
##  $ PartyNm2013              : Factor w/ 9 levels "Australian Labor Party",..: 2 2 2 2 2 2 2 2 2 2 ...

#Manipulate the data to get distinct PC Immun groups

pc_immun_groups <- data %>%
    group_by(pc_immun, pc_immun_class) %>%
    summarize(count = n())

pc_immun_groups

## # A tibble: 9 x 3
## # Groups:   pc_immun [?]
##   pc_immun   pc_immun_class count
##   <fct>               <int> <int>
## 1 <70.0                   1   332
## 2 70.0-74.9               2   612
## 3 75.0-79.9               3  1604
## 4 80.0-84.9               4  5936
## 5 85.0-89.9               5 27468
## 6 90.0-92.4               6 33580
## 7 92.5-94.9               7 35652
## 8 95.0-100.0              8 27160
## 9 NP                      0 58819

#Manipulate the data to get distinct age groups

age_groups <- data %>%
    group_by(age) %>%
    summarize(count = n())

age_groups

## # A tibble: 3 x 2
##   age   count
##   <fct> <int>
## 1 1     63549
## 2 2     63685
## 3 5     63929

Top % coverage areas versus bottom % coverage areas

2. I want to see the bottom postcodes for immunisation coverage and the top postcodes for immunisation coverage

##creating bottom % immunisation coverage in 2016 to group, summarise and arrange
low_pc_immun <- data %>%
  filter(pc_immun_class < 3, pc_immun_class != 0, Index.type == "Index of Relative Socio-economic Advantage and Disadvantage", year == 2016) %>%
  group_by(postcode, PHN_code, state, pc_immun, pc_immun_class, Usual.resident.population) %>%
  summarize(count = n()) %>%
  arrange(pc_immun_class, desc(Usual.resident.population))

low_pc_immun

## # A tibble: 25 x 7
## # Groups:   postcode, PHN_code, state, pc_immun, pc_immun_class [25]
##    postcode PHN_code state pc_immun pc_immun_class Usual.resident.p… count
##       <int> <fct>    <fct> <fct>             <int>             <int> <int>
##  1     2000 PHN101   NSW   <70.0                 1             27411     1
##  2     2481 PHN109   NSW   <70.0                 1             11769     1
##  3     2483 PHN109   NSW   <70.0                 1             10609     1
##  4     4552 PHN306   QLD   <70.0                 1              8432     1
##  5     4102 PHN302   QLD   <70.0                 1              7660     1
##  6     2454 PHN109   NSW   <70.0                 1              7570     1
##  7     2482 PHN109   NSW   <70.0                 1              6280     3
##  8     6083 PHN501   WA    <70.0                 1              3512     1
##  9     6073 PHN501   WA    <70.0                 1              3041     1
## 10     2775 PHN104   NSW   <70.0                 1              1498     3
## # ... with 15 more rows

View(low_pc_immun)

bottom_20_postocdes <- head(low_pc_immun, 20)

bottom_20_postocdes

## # A tibble: 20 x 7
## # Groups:   postcode, PHN_code, state, pc_immun, pc_immun_class [20]
##    postcode PHN_code state pc_immun  pc_immun_class Usual.resident.… count
##       <int> <fct>    <fct> <fct>              <int>            <int> <int>
##  1     2000 PHN101   NSW   <70.0                  1            27411     1
##  2     2481 PHN109   NSW   <70.0                  1            11769     1
##  3     2483 PHN109   NSW   <70.0                  1            10609     1
##  4     4552 PHN306   QLD   <70.0                  1             8432     1
##  5     4102 PHN302   QLD   <70.0                  1             7660     1
##  6     2454 PHN109   NSW   <70.0                  1             7570     1
##  7     2482 PHN109   NSW   <70.0                  1             6280     3
##  8     6083 PHN501   WA    <70.0                  1             3512     1
##  9     6073 PHN501   WA    <70.0                  1             3041     1
## 10     2775 PHN104   NSW   <70.0                  1             1498     3
## 11     3000 PHN201   VIC   70.0-74.9              2            37975     2
## 12     3006 PHN201   VIC   70.0-74.9              2            18808     1
## 13     5000 PHN401   SA    70.0-74.9              2            15115     1
## 14     2481 PHN109   NSW   70.0-74.9              2            11769     2
## 15     2483 PHN109   NSW   70.0-74.9              2            10609     2
## 16     4552 PHN306   QLD   70.0-74.9              2             8432     1
## 17     2454 PHN109   NSW   70.0-74.9              2             7570     1
## 18     6333 PHN503   WA    70.0-74.9              2             5845     1
## 19     6005 PHN501   WA    70.0-74.9              2             5624     2
## 20     2479 PHN109   NSW   70.0-74.9              2             4730     1

#2016 results

#order and rank the top 20 in 2016
pc_immun_ordered <- data %>%
  group_by(postcode, PHN_code, state, pc_immun, pc_immun_class, Usual.resident.population) %>%
  filter(pc_immun_class != 0, Index.type == "Index of Relative Socio-economic Advantage and Disadvantage", year == 2016) %>%
  summarize(count = n()) %>%
  arrange(desc(pc_immun_class), desc(Usual.resident.population))


pc_immun_ordered

## # A tibble: 3,213 x 7
## # Groups:   postcode, PHN_code, state, pc_immun, pc_immun_class [3,213]
##    postcode PHN_code state pc_immun  pc_immun_class Usual.resident.… count
##       <int> <fct>    <fct> <fct>              <int>            <int> <int>
##  1     4350 PHN304   QLD   95.0-100…              8           106745     1
##  2     3977 PHN203   VIC   95.0-100…              8            93326     8
##  3     4740 PHN307   QLD   95.0-100…              8            80089     2
##  4     4670 PHN306   QLD   95.0-100…              8            79606     2
##  5     4305 PHN304   QLD   95.0-100…              8            62549     4
##  6     6065 PHN501   WA    95.0-100…              8            61998     2
##  7     2259 PHN108   NSW   95.0-100…              8            59513     6
##  8     3350 PHN206   VIC   95.0-100…              8            59145     1
##  9     2650 PHN110   NSW   95.0-100…              8            56974     2
## 10     4551 PHN306   QLD   95.0-100…              8            56733     2
## # ... with 3,203 more rows

View(pc_immun_ordered)

top_20_postocdes <- head(pc_immun_ordered, 20)

top_20_postocdes

## # A tibble: 20 x 7
## # Groups:   postcode, PHN_code, state, pc_immun, pc_immun_class [20]
##    postcode PHN_code state  pc_immun pc_immun_class Usual.resident.… count
##       <int> <fct>    <fct>  <fct>             <int>            <int> <int>
##  1     4350 PHN304   QLD    95.0-10…              8           106745     1
##  2     3977 PHN203   VIC    95.0-10…              8            93326     8
##  3     4740 PHN307   QLD    95.0-10…              8            80089     2
##  4     4670 PHN306   QLD    95.0-10…              8            79606     2
##  5     4305 PHN304   QLD    95.0-10…              8            62549     4
##  6     6065 PHN501   WA     95.0-10…              8            61998     2
##  7     2259 PHN108   NSW    95.0-10…              8            59513     6
##  8     3350 PHN206   VIC    95.0-10…              8            59145     1
##  9     2650 PHN110   NSW    95.0-10…              8            56974     2
## 10     4551 PHN306   QLD    95.0-10…              8            56733     2
## 11     3216 PHN206   VIC    95.0-10…              8            56284     1
## 12     4300 PHN304   QLD    95.0-10…              8            55669     2
## 13     2155 PHN103   NSW    95.0-10…              8            55452     2
## 14     4680 PHN306   QLD    95.0-10…              8            54166     1
## 15     2261 PHN108   NSW    95.0-10…              8            52754     2
## 16     3037 PHN201   VIC    95.0-10…              8            51155     2
## 17     2340 PHN108   NSW    95.0-10…              8            49072     4
## 18     3810 PHN203   VIC    95.0-10…              8            47894     2
## 19     2444 PHN109   NSW    95.0-10…              8            46284     2
## 20     2540 PHN106   NSW/A… 95.0-10…              8            45121     4

bottom_20_postocdes <- tail(pc_immun_ordered, 20)

arrange(bottom_20_postocdes, pc_immun_class, desc(Usual.resident.population))

## # A tibble: 20 x 7
## # Groups:   postcode, PHN_code, state, pc_immun, pc_immun_class [20]
##    postcode PHN_code state pc_immun  pc_immun_class Usual.resident.… count
##       <int> <fct>    <fct> <fct>              <int>            <int> <int>
##  1     2000 PHN101   NSW   <70.0                  1            27411     1
##  2     2481 PHN109   NSW   <70.0                  1            11769     1
##  3     2483 PHN109   NSW   <70.0                  1            10609     1
##  4     4552 PHN306   QLD   <70.0                  1             8432     1
##  5     4102 PHN302   QLD   <70.0                  1             7660     1
##  6     2454 PHN109   NSW   <70.0                  1             7570     1
##  7     2482 PHN109   NSW   <70.0                  1             6280     3
##  8     6083 PHN501   WA    <70.0                  1             3512     1
##  9     6073 PHN501   WA    <70.0                  1             3041     1
## 10     2775 PHN104   NSW   <70.0                  1             1498     3
## 11     4552 PHN306   QLD   70.0-74.9              2             8432     1
## 12     2454 PHN109   NSW   70.0-74.9              2             7570     1
## 13     6333 PHN503   WA    70.0-74.9              2             5845     1
## 14     6005 PHN501   WA    70.0-74.9              2             5624     2
## 15     2479 PHN109   NSW   70.0-74.9              2             4730     1
## 16     3461 PHN206   VIC   70.0-74.9              2             4149     1
## 17     2469 PHN109   NSW   70.0-74.9              2             4012     2
## 18     5172 PHN401   SA    70.0-74.9              2             3932     2
## 19     4677 PHN306   QLD   70.0-74.9              2             3549     1
## 20     2848 PHN107   NSW   70.0-74.9              2             1645     1

3. I want to know the mean immunisation % coverage rate across Australia in 2011 - but I also want to know what the mean immunisation% coverage rate is for each state.

#filter and explore the 2011 data
mean_immun_rate_aust_2011 <- data %>%
    filter(year == 2011, pc_immun_class != 0)

mean(mean_immun_rate_aust_2011$pc_immun_class)

## [1] 6.271792

#filter, group and summarise the 2011 data by state
mean_immun_rate_state_2011 <- data %>%
    filter(year == 2011) %>%
    group_by(state) %>%
    summarize(
        mean = ifelse(sum(pc_immun_class[pc_immun_class != 0]) > 0, mean(pc_immun_class[pc_immun_class != 0]), 0),
        rounded_mean = round(as.numeric(mean)),
        pc_immun = pc_immun_groups$pc_immun[pc_immun_groups$pc_immun_class == rounded_mean],
        total_rows = n(),
        num_of_non_NP = sum(pc_immun_class != 0),
        num_of_NP = sum(pc_immun_class == 0),
        proportion_of_NP = (num_of_NP/n()))

mean_immun_rate_state_2011

## # A tibble: 8 x 8
##   state  mean rounded_mean pc_immun total_rows num_of_non_NP num_of_NP
##   <fct> <dbl>        <dbl> <fct>         <int>         <int>     <int>
## 1 ACT    6.38            6 90.0-92…        316           316         0
## 2 NSW    6.28            6 90.0-92…       6920          6920         0
## 3 NT     6.71            7 92.5-94…        220           220         0
## 4 QLD    6.28            6 90.0-92…       4268          4268         0
## 5 SA     6.03            6 90.0-92…       2252          2252         0
## 6 TAS    6.53            7 92.5-94…        772           772         0
## 7 VIC    6.56            7 92.5-94…       5308          5308         0
## 8 WA     5.61            6 90.0-92…       2108          2108         0
## # ... with 1 more variable: proportion_of_NP <dbl>

4. I want to know the mean immunisation % coverage rate across Australia in 2016 - and I want to know what the mean is for each state (even those weird doubly states we have)

mean_immun_rate_aust_2016 <- data %>%
    filter(year == 2016, pc_immun_class != 0)

mean(mean_immun_rate_aust_2016$pc_immun_class)

## [1] 6.63243

#filter, group and summarise the 2016 data by state
mean_immun_rate_state_2016 <- data %>%
    filter(year == 2016) %>%
    group_by(state) %>%
    summarize(
        mean = ifelse(sum(pc_immun_class[pc_immun_class != 0]) > 0, mean(pc_immun_class[pc_immun_class != 0]), 0),
        rounded_mean = round(as.numeric(mean)),
        pc_immun = pc_immun_groups$pc_immun[pc_immun_groups$pc_immun_class == rounded_mean],
        total_rows = n(),
        num_of_non_NP = sum(pc_immun_class != 0),
        num_of_NP = sum(pc_immun_class == 0),
        proportion_of_NP = (num_of_NP/n()))


mean_immun_rate_state_2016

## # A tibble: 12 x 8
##    state  mean rounded_mean pc_immun total_rows num_of_non_NP num_of_NP
##    <fct> <dbl>        <dbl> <fct>         <int>         <int>     <int>
##  1 ACT    6.64            7 92.5-94…        312           280        32
##  2 NSW    6.58            7 92.5-94…       8648          6800      1848
##  3 NSW/…  7.52            8 95.0-10…        108            84        24
##  4 NSW/…  0               0 NP               48             0        48
##  5 NSW/…  7               7 92.5-94…         24            24         0
##  6 NT     6.22            6 90.0-92…        324           180       144
##  7 QLD    6.82            7 92.5-94…       5919          4180      1739
##  8 SA     6.61            7 92.5-94…       4232          2236      1996
##  9 SA/W…  5.67            6 90.0-92…         24            24         0
## 10 TAS    6.82            7 92.5-94…       1440           704       736
## 11 VIC    6.76            7 92.5-94…       9176          5364      3812
## 12 WA     6.10            6 90.0-92…       4072          2128      1944
## # ... with 1 more variable: proportion_of_NP <dbl>

View(mean_immun_rate_state_2016)

5. I want to see meaningful electorate and party in power for 2011 and 2016 - I don’t know whether this would just be a smaller dataset or if we can somehow plot it???

#Get info by electorate for 2016
electorates <- data %>%
    filter(Index.type == "Index of Relative Socio-economic Advantage and Disadvantage") %>%
    group_by(Electoral.division, PartyNm2016, year) %>%
    summarize(
        mean = ifelse(sum(pc_immun_class[pc_immun_class != 0]) > 0, mean(pc_immun_class[pc_immun_class != 0]), 0),
        rounded_mean = round(as.numeric(mean)),
        pc_immun = pc_immun_groups$pc_immun[pc_immun_groups$pc_immun_class == rounded_mean],
        total_rows = n(),
        num_of_non_NP = sum(pc_immun_class != 0),
        num_of_NP = sum(pc_immun_class == 0),
        proportion_of_NP = (num_of_NP/n())) %>%
    arrange(mean)

Get the electorates with the lowest 10 mean pc_immun_class scores for 2016

lowest_pc_immun <- head(filter(electorates, year == 2016), 10)

lowest_pc_immun

## # A tibble: 10 x 10
## # Groups:   Electoral.division, PartyNm2016 [10]
##    Electoral.divis… PartyNm2016  year  mean rounded_mean pc_immun
##    <fct>            <fct>       <int> <dbl>        <dbl> <fct>   
##  1 Richmond         Australian…  2016  4.38            4 80.0-84…
##  2 Fairfax          Liberal Na…  2016  5.26            5 85.0-89…
##  3 Wentworth        Liberal      2016  5.36            5 85.0-89…
##  4 Curtin           Liberal      2016  5.56            6 90.0-92…
##  5 Reid             Liberal      2016  5.61            6 90.0-92…
##  6 Swan             Liberal      2016  5.63            6 90.0-92…
##  7 Perth            Australian…  2016  5.65            6 90.0-92…
##  8 Cowper           The Nation…  2016  5.68            6 90.0-92…
##  9 Fisher           Liberal Na…  2016  5.69            6 90.0-92…
## 10 Fremantle        Australian…  2016  5.70            6 90.0-92…
## # ... with 4 more variables: total_rows <int>, num_of_non_NP <int>,
## #   num_of_NP <int>, proportion_of_NP <dbl>

Get those electorate names

lowest_pc_immun_names <- select(lowest_pc_immun, Electoral.division, PartyNm2016)
lowest_pc_immun_names <- as.character(lowest_pc_immun$Electoral.division)

Let’s look at postcodes in the electorates with lowest mean pc_immun_class scores

#Get postciode level data for the lowest10 electorates in 2016
lowest_electorates_2016 <- data %>%
    filter(Electoral.division %in% lowest_pc_immun_names) %>%
    select(state, year, Index.type, Electoral.division, postcode, age, pc_immun, pc_immun_class, PHN_code, Score)

ggplot(filter(lowest_electorates_2016, Index.type == "Index of Relative Socio-economic Disadvantage"), aes(age)) +
  geom_point(aes(y = pc_immun, size=Score, color = age), na.rm = TRUE) +
  facet_wrap(~ year, ncol = 3, nrow = 5) +
  labs(title="Relative Socio-economic Disadvantage Scores", subtitle = "Bottom 10 electorates in 2016: a retrospective")

#OK. Let's look at postcodes in those electorates
#Get info by electorate
ggplot(filter(lowest_electorates_2016, Index.type == "Index of Relative Socio-economic Advantage and Disadvantage"), aes(age)) +
  geom_point(aes(y = pc_immun, size=Score, color = age), na.rm = TRUE) +
  facet_wrap(~ year, ncol = 3) +
  labs(title="Relative Socio-economic Advantage and Disadvantage Scores", subtitle = "Bottom 10 electorates in 2016: a retrospective")

#OK. Let's look at postcodes in those electorates
#Get info by electorate
ggplot(filter(lowest_electorates_2016, Index.type == "Index of Economic Resources"), aes(age)) +
  geom_point(aes(y = pc_immun, size=Score, color = age), na.rm = TRUE) +
  facet_wrap(~ year, ncol = 3) +
  labs(title="Economic Resources Scores", subtitle = "Bottom 10 electorates in 2016: a retrospective")

#OK. Let's look at postcodes in those electorates
#Get info by electorate
ggplot(filter(lowest_electorates_2016, Index.type == "Index of Education and Occupation"), aes(age)) +
  geom_point(aes(y = pc_immun, size=Score, color = age), na.rm = TRUE) +
  facet_wrap(~ year, ncol = 3) +
  labs(title="Education and Occupation Scores", subtitle = "Bottom 10 electorates in 2016: a retrospective")

highest_pc_immun <- tail(filter(electorates, year == 2016), 10)

highest_pc_immun

## # A tibble: 10 x 10
## # Groups:   Electoral.division, PartyNm2016 [10]
##    Electoral.divis… PartyNm2016  year  mean rounded_mean pc_immun
##    <fct>            <fct>       <int> <dbl>        <dbl> <fct>   
##  1 Charlton         <NA>         2016  7.28            7 92.5-94…
##  2 Herbert          Australian…  2016  7.29            7 92.5-94…
##  3 Macarthur        Australian…  2016  7.29            7 92.5-94…
##  4 Lilley           Australian…  2016  7.30            7 92.5-94…
##  5 Dawson           Liberal Na…  2016  7.34            7 92.5-94…
##  6 Riverina         The Nation…  2016  7.38            7 92.5-94…
##  7 Wannon           Liberal      2016  7.42            7 92.5-94…
##  8 Hunter           Australian…  2016  7.42            7 92.5-94…
##  9 Hughes           Liberal      2016  7.46            7 92.5-94…
## 10 Shortland        Australian…  2016  7.78            8 95.0-10…
## # ... with 4 more variables: total_rows <int>, num_of_non_NP <int>,
## #   num_of_NP <int>, proportion_of_NP <dbl>

highest_pc_immun_names <- select(highest_pc_immun, Electoral.division, PartyNm2016)
highest_pc_immun_names <- as.character(highest_pc_immun$Electoral.division)

#OK. Let's look at postcodes in those electorates
#Get info by electorate
#Get postcode level data for the highest 10 electorates in 2016
highest_electorates_2016 <- data %>%
    filter(Electoral.division %in% highest_pc_immun_names) %>%
    select(state, year, Electoral.division, Index.type, postcode, age, pc_immun, pc_immun_class, PHN_code, Score)

ggplot(filter(highest_electorates_2016, Index.type == "Index of Relative Socio-economic Disadvantage"), aes(age)) +
  geom_point(aes(y = pc_immun, size=Score, color = age), na.rm = TRUE) +
  facet_wrap(~ year, ncol = 3) +
  labs(title="Relative Socio-economic Disadvantage Score", subtitle = "Top 10 highest electorates in 2016: a retrospective")

#OK. Let's look at postcodes in those electorates
#Get info by electorate
ggplot(filter(highest_electorates_2016, Index.type == "Index of Relative Socio-economic Advantage and Disadvantage"), aes(age)) +
  geom_point(aes(y = pc_immun, size=Score, color = age), na.rm = TRUE) +
  facet_wrap(~ year, ncol = 3) +
  labs(title="Relative Socio-economic Advantage and Disadvantage Scores", subtitle = "Top 10 highest electorates in 2016: a retrospective")

#OK. Let's look at postcodes in those electorates
#Get info by electorate
ggplot(filter(highest_electorates_2016, Index.type == "Index of Economic Resources"), aes(age)) +
  geom_point(aes(y = pc_immun, size=Score, color = age), na.rm = TRUE) +
  facet_wrap(~ year, ncol = 3) +
  labs(title="Economic Resources Scores", subtitle = "Top 10 highest electorates in 2016: a retrospective")

ggplot(filter(highest_electorates_2016, Index.type == "Index of Education and Occupation"), aes(age)) +
  geom_point(aes(y = pc_immun, size = Score, color = age), na.rm = TRUE) +
  facet_wrap(~ year, ncol = 3) +
  labs(title="Education and Occupation Scores", subtitle = "Top 10 highest electorates in 2016: a retrospective")

#Get info by 2016political party
party_2016 <- data %>%
    group_by(PartyNm2016) %>%
    summarize(
        mean = ifelse(sum(pc_immun_class[pc_immun_class != 0]) > 0, mean(pc_immun_class[pc_immun_class != 0]), 0),
        rounded_mean = round(as.numeric(mean)),
        pc_immun = pc_immun_groups$pc_immun[pc_immun_groups$pc_immun_class == rounded_mean],
        total_rows = n(),
        num_of_non_NP = sum(pc_immun_class != 0),
        num_of_NP = sum(pc_immun_class == 0),
        proportion_of_NP = (num_of_NP/n()))

party_2016

## # A tibble: 9 x 8
##   PartyNm2016  mean rounded_mean pc_immun total_rows num_of_non_NP
##   <fct>       <dbl>        <dbl> <fct>         <int>         <int>
## 1 Australian…  6.27            6 90.0-92…      65980         54932
## 2 Independent  6.68            7 92.5-94…       3864          1636
## 3 Katter's A…  6.69            7 92.5-94…       2296          1528
## 4 Liberal      6.24            6 90.0-92…      61084         38884
## 5 Liberal Na…  6.37            6 90.0-92…      21935         15824
## 6 Nick Xenop…  5.74            6 90.0-92…       3464          1476
## 7 The Greens   5.85            6 90.0-92…       1280          1224
## 8 The Nation…  6.58            7 92.5-94…      19584          9024
## 9 <NA>         6.60            7 92.5-94…      11676          7816
## # ... with 2 more variables: num_of_NP <int>, proportion_of_NP <dbl>

6. I want to somehow try to map or plot the SEIFA scores for the bottom 20 and top 20 postcodes for immunisation coverage - again, this might be too ambitious

We know lots of postcodes are missing % immunisation coverage

There seems to be some missing electoral data too

#check for missing values
View(data)

#na_data <- data %>%
    #summarize(
      #state_na = count(is.na(state)),
      #state_total = count(state),
    #)

#na_data

missmap(data, main = "Missing values vs observed")

#read this to see if I should impute missing data https://www.linkedin.com/pulse/amelia-packager-missing-data-imputation-ramprakash-veluchamy/

#try plotting postcode to immun_class over the years - not very visual
ggplot(data = data, mapping = aes(x = postcode,y = pc_immun_class, color = year))+
         geom_point()

# Get % immunised counts per pc immun class for each year
counts_by_year_merged<-data %>%
  group_by(year, pc_immun_class) %>%
  summarize(count = n())

counts_by_year_merged$pc_immun = NA
View(counts_by_year_merged)

Now try plots

#plot how % immunised has changed by category between 2011 to 2016 - I would like to not have the bottom counts so cluttered on this axis
ggplot(data = counts_by_year_merged) +
  geom_point(mapping = aes(x=count, y=year, color=pc_immun_class))+
  facet_wrap(~pc_immun, nrow = 2)