Zombies Apocalypse

• Data Source: Kaggle
• Tasks: compare humans and zombies to identify differences in supplies
• Language: julia

Context

News reports suggest that the impossible has become possible…zombies have appeared on the streets of the US! What should we do? The Centers for Disease Control and Prevention (CDC) zombie preparedness website recommends storing water, food, medication, tools, sanitation items, clothing, essential documents, and first aid supplies. Thankfully, we are CDC analysts and are prepared, but it may be too late for others!

Content

Our team decides to identify supplies that protect people and coordinate supply distribution. A few brave data collectors volunteer to check on 200 randomly selected adults who were alive before the zombies. We have recent data for the 200 on age and sex, how many are in their household, and their rural, suburban, or urban location. Our heroic volunteers visit each home and record zombie status and preparedness. Now it's our job to figure out which supplies are associated with safety!

File

Because every moment counts when dealing with life and (un)death, we want to get this right! The first task is to compare humans and zombies to identify differences in supplies. We review the data and find the following:

• zombieid: unique identifier
• zombie: human or zombie
• age: age in years
• sex: male or female
• rurality: rural, suburban, or urban
• household: number of people living in household
• water: gallons of clean water available
• food: food or no food
• medication: medication or no medication
• tools: tools or no tools
• firstaid: first aid or no first aid
• sanitation: sanitation or no sanitation
• clothing: clothing or no clothing
• documents: documents or no documents

Acknowledgements

DataCamp

ENV["COLUMNS"] = 1000; # print more columns of tables

using Random

Random.seed!(42)
"Andi Kerstin Chris Caro Jana" |> split |> shuffle |> x -> join(x," → ")

"Kerstin → Chris → Andi → Jana → Caro"

1. Data loading

using Dates
using CSV
using DataFrames

data = CSV.read("zombies.csv", DataFrame)
first(data, 5)

5×14 DataFrame

Row	zombieid	zombie	age	sex	rurality	household	water	food	medication	tools	firstaid	sanitation	clothing	documents
	Int64	String7	Int64	String7	String15	Int64	Int64	String7	String15	String15	String31	String15	String15	String15
1	1	Human	18	Female	Rural	1	0	Food	Medication	No tools	First aid supplies	Sanitation	Clothing	NA
2	2	Human	18	Male	Rural	3	24	Food	Medication	tools	First aid supplies	Sanitation	Clothing	NA
3	3	Human	18	Male	Rural	4	16	Food	Medication	No tools	First aid supplies	Sanitation	Clothing	NA
4	4	Human	19	Male	Rural	1	0	Food	Medication	tools	No first aid supplies	Sanitation	Clothing	NA
5	5	Human	19	Male	Urban	1	0	Food	Medication	No tools	First aid supplies	Sanitation	NA	NA

unique(data.food)

2-element Vector{String7}:
 "Food"
 "No food"

unique(data.medication)

2-element Vector{String15}:
 "Medication"
 "No medication"

unique(data.tools)

2-element Vector{String15}:
 "No tools"
 "tools"

unique(data.firstaid)

2-element Vector{String31}:
 "First aid supplies"
 "No first aid supplies"

unique(data.sanitation)

2-element Vector{String15}:
 "Sanitation"
 "No sanitation"

unique(data.clothing)

2-element Vector{String15}:
 "Clothing"
 "NA"

unique(data.documents)

2-element Vector{String15}:
 "NA"
 "Documents"

using GLMakie

using DataFramesMeta
using Chain
using StatsBase

unique(data.age)
dict_age = sort(countmap(data.age))

OrderedCollections.OrderedDict{Int64, Int64} with 62 entries:
  18 => 4
  19 => 4
  20 => 3
  21 => 5
  22 => 1
  23 => 4
  24 => 5
  25 => 7
  26 => 4
  27 => 2
  28 => 6
  29 => 6
  30 => 4
  31 => 2
  32 => 8
  33 => 3
  34 => 2
  35 => 2
  36 => 5
  ⋮  => ⋮

collect(values(dict_age))

62-element Vector{Int64}:
 4
 4
 3
 5
 1
 4
 5
 7
 4
 2
 ⋮
 1
 2
 3
 1
 2
 1
 2
 1
 1

f, ax, plt = hist(data_grouped[1].age, color = (:blue, 0.5), label = "Human")
hist!(ax, data_grouped[2].age, color = (:red, 0.5), label = "Zombie")
axislegend(ax)
display(f)

GLMakie.Screen(...)

data_grouped = groupby(data, :zombie)

GroupedDataFrame with 2 groups based on key: zombie

First Group (121 rows): zombie = "Human"

96 rows omitted

Row	zombieid	zombie	age	sex	rurality	household	water	food	medication	tools	firstaid	sanitation	clothing	documents
	Int64	String7	Int64	String7	String15	Int64	Int64	String7	String15	String15	String31	String15	String15	String15
1	1	Human	18	Female	Rural	1	0	Food	Medication	No tools	First aid supplies	Sanitation	Clothing	NA
2	2	Human	18	Male	Rural	3	24	Food	Medication	tools	First aid supplies	Sanitation	Clothing	NA
3	3	Human	18	Male	Rural	4	16	Food	Medication	No tools	First aid supplies	Sanitation	Clothing	NA
4	4	Human	19	Male	Rural	1	0	Food	Medication	tools	No first aid supplies	Sanitation	Clothing	NA
5	5	Human	19	Male	Urban	1	0	Food	Medication	No tools	First aid supplies	Sanitation	NA	NA
6	6	Human	19	Female	Urban	1	0	Food	Medication	tools	First aid supplies	Sanitation	Clothing	NA
7	7	Human	20	Female	Suburban	2	0	No food	Medication	No tools	First aid supplies	Sanitation	Clothing	NA
8	8	Human	20	Female	Rural	2	0	Food	No medication	No tools	No first aid supplies	Sanitation	Clothing	NA
9	9	Human	21	Female	Urban	1	8	No food	No medication	tools	First aid supplies	Sanitation	Clothing	Documents
10	10	Human	21	Female	Rural	2	8	No food	No medication	tools	First aid supplies	Sanitation	Clothing	Documents
11	11	Human	21	Male	Rural	1	8	Food	No medication	No tools	First aid supplies	No sanitation	NA	NA
12	12	Human	21	Male	Rural	2	16	No food	Medication	No tools	No first aid supplies	Sanitation	Clothing	Documents
13	13	Human	22	Male	Suburban	2	16	Food	Medication	No tools	First aid supplies	Sanitation	Clothing	Documents
⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮
110	110	Human	63	Female	Rural	1	0	Food	Medication	No tools	First aid supplies	Sanitation	Clothing	Documents
111	111	Human	65	Female	Rural	2	16	Food	No medication	tools	No first aid supplies	No sanitation	NA	NA
112	112	Human	67	Male	Rural	2	16	No food	No medication	tools	No first aid supplies	No sanitation	NA	NA
113	113	Human	68	Male	Rural	2	8	No food	Medication	No tools	First aid supplies	Sanitation	Clothing	Documents
114	114	Human	69	Female	Rural	2	8	Food	Medication	No tools	First aid supplies	No sanitation	NA	NA
115	115	Human	71	Female	Urban	2	8	Food	Medication	No tools	No first aid supplies	Sanitation	Clothing	Documents
116	116	Human	72	Male	Suburban	2	0	Food	Medication	tools	No first aid supplies	No sanitation	Clothing	NA
117	117	Human	74	Male	Suburban	1	0	Food	Medication	tools	First aid supplies	No sanitation	Clothing	NA
118	118	Human	75	Female	Rural	1	8	Food	No medication	tools	First aid supplies	No sanitation	NA	NA
119	119	Human	77	Female	Rural	1	8	Food	Medication	No tools	No first aid supplies	No sanitation	NA	NA
120	120	Human	81	Male	Rural	1	8	Food	Medication	tools	No first aid supplies	No sanitation	NA	NA
121	121	Human	32	Male	Rural	2	8	Food	No medication	No tools	First aid supplies	Sanitation	Clothing	Documents

⋮

Last Group (79 rows): zombie = "Zombie"

54 rows omitted

Row	zombieid	zombie	age	sex	rurality	household	water	food	medication	tools	firstaid	sanitation	clothing	documents
	Int64	String7	Int64	String7	String15	Int64	Int64	String7	String15	String15	String31	String15	String15	String15
1	122	Zombie	20	Female	Urban	2	0	Food	No medication	tools	First aid supplies	No sanitation	Clothing	NA
2	123	Zombie	23	Male	Suburban	3	0	No food	No medication	No tools	No first aid supplies	No sanitation	Clothing	NA
3	124	Zombie	25	Female	Rural	5	0	No food	No medication	No tools	No first aid supplies	No sanitation	Clothing	NA
4	125	Zombie	28	Female	Suburban	3	0	No food	Medication	tools	First aid supplies	No sanitation	Clothing	NA
5	126	Zombie	31	Female	Rural	4	0	No food	No medication	tools	First aid supplies	No sanitation	Clothing	NA
6	127	Zombie	32	Male	Suburban	4	0	No food	No medication	No tools	No first aid supplies	Sanitation	NA	Documents
7	128	Zombie	42	Male	Rural	4	8	No food	No medication	No tools	No first aid supplies	Sanitation	Clothing	Documents
8	129	Zombie	43	Male	Urban	5	8	No food	No medication	tools	First aid supplies	No sanitation	Clothing	NA
9	130	Zombie	44	Male	Rural	5	8	Food	No medication	tools	First aid supplies	No sanitation	Clothing	NA
10	131	Zombie	45	Male	Urban	4	0	Food	Medication	No tools	No first aid supplies	No sanitation	Clothing	NA
11	132	Zombie	47	Female	Urban	2	0	No food	Medication	No tools	No first aid supplies	No sanitation	Clothing	NA
12	133	Zombie	48	Female	Suburban	3	0	No food	No medication	No tools	First aid supplies	No sanitation	Clothing	NA
13	134	Zombie	48	Female	Urban	2	0	No food	No medication	tools	First aid supplies	Sanitation	Clothing	Documents
⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮	⋮
68	189	Zombie	32	Male	Urban	3	0	No food	No medication	No tools	First aid supplies	No sanitation	Clothing	NA
69	190	Zombie	41	Female	Rural	5	0	No food	No medication	tools	First aid supplies	No sanitation	Clothing	NA
70	191	Zombie	43	Female	Rural	5	0	No food	No medication	No tools	No first aid supplies	Sanitation	Clothing	Documents
71	192	Zombie	48	Female	Suburban	4	8	No food	No medication	No tools	No first aid supplies	No sanitation	NA	NA
72	193	Zombie	58	Male	Urban	1	0	Food	No medication	tools	First aid supplies	No sanitation	NA	NA
73	194	Zombie	65	Male	Urban	1	0	No food	No medication	tools	First aid supplies	No sanitation	NA	NA
74	195	Zombie	67	Female	Suburban	2	0	No food	No medication	No tools	No first aid supplies	No sanitation	NA	NA
75	196	Zombie	68	Male	Suburban	1	0	Food	No medication	No tools	No first aid supplies	Sanitation	Clothing	Documents
76	197	Zombie	71	Male	Suburban	1	8	No food	No medication	tools	First aid supplies	No sanitation	Clothing	NA
77	198	Zombie	76	Female	Urban	1	0	No food	No medication	tools	First aid supplies	Sanitation	Clothing	Documents
78	199	Zombie	82	Male	Urban	1	0	No food	No medication	No tools	No first aid supplies	No sanitation	NA	NA
79	200	Zombie	85	Male	Urban	1	0	No food	Medication	No tools	No first aid supplies	Sanitation	Clothing	NA

countmap(data_grouped[1].sex)

Dict{String7, Int64} with 2 entries:
  "Female" => 62
  "Male"   => 59

first(data)

DataFrameRow (14 columns)

Row	zombieid	zombie	age	sex	rurality	household	water	food	medication	tools	firstaid	sanitation	clothing	documents
	Int64	String7	Int64	String7	String15	Int64	Int64	String7	String15	String15	String31	String15	String15	String15
1	1	Human	18	Female	Rural	1	0	Food	Medication	No tools	First aid supplies	Sanitation	Clothing	NA

data_count = @chain data begin
    groupby(:zombie)
    @combine(:sex = countmap(:sex), 
    :rurality = countmap(:rurality), 
    :food = countmap(:food), 
    :medication = countmap(:medication), 
    :tools = countmap(:tools), 
    :firstaid = countmap(:firstaid), 
    :sanitation = countmap(:sanitation), 
    :clothing = countmap(:clothing), 
    :documents = countmap(:documents))
end

2×10 DataFrame

Row	zombie	sex	rurality	food	medication	tools	firstaid	sanitation	clothing	documents
	String7	Dict…	Dict…	Dict…	Dict…	Dict…	Dict…	Dict…	Dict…	Dict…
1	Human	Dict{String7, Int64}("Female"=>62, "Male"=>59)	Dict{String15, Int64}("Urban"=>16, "Rural"=>80, "Suburban"=>25)	Dict{String7, Int64}("Food"=>91, "No food"=>30)	Dict{String15, Int64}("No medication"=>43, "Medication"=>78)	Dict{String15, Int64}("tools"=>60, "No tools"=>61)	Dict{String31, Int64}("First aid supplies"=>67, "No first aid supplies"=>54)	Dict{String15, Int64}("Sanitation"=>73, "No sanitation"=>48)	Dict{String15, Int64}("NA"=>47, "Clothing"=>74)	Dict{String15, Int64}("NA"=>77, "Documents"=>44)
2	Zombie	Dict{String7, Int64}("Female"=>37, "Male"=>42)	Dict{String15, Int64}("Urban"=>38, "Rural"=>18, "Suburban"=>23)	Dict{String7, Int64}("Food"=>19, "No food"=>60)	Dict{String15, Int64}("No medication"=>63, "Medication"=>16)	Dict{String15, Int64}("tools"=>39, "No tools"=>40)	Dict{String31, Int64}("First aid supplies"=>39, "No first aid supplies"=>40)	Dict{String15, Int64}("Sanitation"=>25, "No sanitation"=>54)	Dict{String15, Int64}("NA"=>27, "Clothing"=>52)	Dict{String15, Int64}("NA"=>57, "Documents"=>22)

colors = [:red, :blue]
elem_1 = [PolyElement(color = :red, strokecolor = :blue, strokewidth = 1)]
elem_2 = [PolyElement(color = :blue, strokecolor = :blue, strokewidth = 1)]

1-element Vector{PolyElement}:
 PolyElement(Attributes with 3 entries:
  polycolor => blue
  polystrokecolor => blue
  polystrokewidth => 1)

f, ax, plt = pie(collect(values(data_count.sex[1])),
                 color = colors,
                 radius = 4,
                 inner_radius = 2,
                 strokecolor = :white,
                 strokewidth = 5,
                 axis = ( autolimitaspect = 1, ))
ax2 = Axis(f[1,2],  autolimitaspect = 1, )
pie!(ax2, collect(values(data_count.sex[2])),
color = colors,
radius = 4,
inner_radius = 2,
strokecolor = :white,
strokewidth = 5)
Legend(f[1, 3],
    [elem_1, elem_2],
    ["Female", "Male"],
    patchsize = (35, 35), rowgap = 10)
display(f)

GLMakie.Screen(...)

f, ax, plt = pie(collect(values(data_count.food[1])),
                 color = colors,
                 radius = 4,
                 inner_radius = 2,
                 strokecolor = :white,
                 strokewidth = 5,
                 axis = ( autolimitaspect = 1, ))
                
ax2 = Axis(f[1,2], autolimitaspect = 1, )
pie!(ax2, collect(values(data_count.food[2])),
color = colors,
radius = 4,
inner_radius = 2,
strokecolor = :white,
strokewidth = 5)
Legend(f[1, 3],
    [elem_1, elem_2],
    ["Food", "No Food"],
    patchsize = (35, 35), rowgap = 10)
display(f)

GLMakie.Screen(...)

f, ax, plt = pie(collect(values(data_count.medication[1])),
                 color = colors,
                 radius = 4,
                 inner_radius = 2,
                 strokecolor = :white,
                 strokewidth = 5,
                 axis = ( autolimitaspect = 1, ))
                
ax2 = Axis(f[1,2] , autolimitaspect = 1, )
pie!(ax2, collect(values(data_count.medication[2])),
color = colors,
radius = 4,
inner_radius = 2,
strokecolor = :white,
strokewidth = 5)
Legend(f[1, 3],
    [elem_1, elem_2],
    ["Medication", "No Medication"],
    patchsize = (35, 35), rowgap = 10)
display(f)

GLMakie.Screen(...)