STAT 170 Project EDA - Rhea Verma
2025-11-07
Question: Is there a difference in life expectancy between developed and undeveloped countries? (Compare other variables across developed countries vs undeveloped countries.)
- Dataset with Status, Life expectancy, Education, Adult Mortality, Infant deaths, Population
- Y - Life expectancy
- X (Main predictor) - Status (Developing Vs Developed)
- Status is binary (Developing Vs Developed)
- Sub Question: How do different factors help determine/predict if a country is developing vs developed?
- Include other predictors - Education, Adult Mortality, Infant deaths, Population
Import Libraries
library(tidyverse)## ── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
## ✔ dplyr 1.1.4 ✔ readr 2.1.5
## ✔ forcats 1.0.1 ✔ stringr 1.6.0
## ✔ ggplot2 4.0.0 ✔ tibble 3.3.0
## ✔ lubridate 1.9.4 ✔ tidyr 1.3.1
## ✔ purrr 1.2.0
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag() masks stats::lag()
## ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errorslibrary(ggplot2)
library(dplyr)
library(corrplot)## corrplot 0.95 loadedlibrary(GGally)Step 1: Random Split Data (70/30)
LE <- na.omit(read.csv("Life Expectancy Data.csv"))
set.seed(127)
n <- nrow(LE)
train_index <- sample(1:n, size = 0.7*n)
train_data <- LE[train_index, ]
test_data <- LE[-train_index, ]Step 2: Data Description
Dataset source and context:
- The data used in this study come from the World Health Organization (WHO) and report life expectancy and related health, demographic, and economic indicators for countries worldwide from 2000 to 2015 with 2938 observations (before the cleaning of the dataset). The dataset provides a cross-section of public health conditions across both Developed and Developing countries. Each row represents a specific country-year observation.
Definitions, types, and units of all variables:
str(LE)## 'data.frame': 1649 obs. of 22 variables:
## $ Country : chr "Afghanistan" "Afghanistan" "Afghanistan" "Afghanistan" ...
## $ Year : int 2015 2014 2013 2012 2011 2010 2009 2008 2007 2006 ...
## $ Status : chr "Developing" "Developing" "Developing" "Developing" ...
## $ Life.expectancy : num 65 59.9 59.9 59.5 59.2 58.8 58.6 58.1 57.5 57.3 ...
## $ Adult.Mortality : int 263 271 268 272 275 279 281 287 295 295 ...
## $ infant.deaths : int 62 64 66 69 71 74 77 80 82 84 ...
## $ Alcohol : num 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.03 0.02 0.03 ...
## $ percentage.expenditure : num 71.3 73.5 73.2 78.2 7.1 ...
## $ Hepatitis.B : int 65 62 64 67 68 66 63 64 63 64 ...
## $ Measles : int 1154 492 430 2787 3013 1989 2861 1599 1141 1990 ...
## $ BMI : num 19.1 18.6 18.1 17.6 17.2 16.7 16.2 15.7 15.2 14.7 ...
## $ under.five.deaths : int 83 86 89 93 97 102 106 110 113 116 ...
## $ Polio : int 6 58 62 67 68 66 63 64 63 58 ...
## $ Total.expenditure : num 8.16 8.18 8.13 8.52 7.87 9.2 9.42 8.33 6.73 7.43 ...
## $ Diphtheria : int 65 62 64 67 68 66 63 64 63 58 ...
## $ HIV.AIDS : num 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 ...
## $ GDP : num 584.3 612.7 631.7 670 63.5 ...
## $ Population : num 33736494 327582 31731688 3696958 2978599 ...
## $ thinness..1.19.years : num 17.2 17.5 17.7 17.9 18.2 18.4 18.6 18.8 19 19.2 ...
## $ thinness.5.9.years : num 17.3 17.5 17.7 18 18.2 18.4 18.7 18.9 19.1 19.3 ...
## $ Income.composition.of.resources: num 0.479 0.476 0.47 0.463 0.454 0.448 0.434 0.433 0.415 0.405 ...
## $ Schooling : num 10.1 10 9.9 9.8 9.5 9.2 8.9 8.7 8.4 8.1 ...
## - attr(*, "na.action")= 'omit' Named int [1:1289] 33 45 46 47 48 49 58 59 60 61 ...
## ..- attr(*, "names")= chr [1:1289] "33" "45" "46" "47" ...- Country (Categorical): Name of the country.
- Year (Numerical/Integer): Year observed.
- Status (Categorical/Binary): Developed or Developing status.
- Life.expectancy (Numerical): Average expected lifespan at birth in age.
- Adult.Mortality (Numerical/Integer): Adult Mortality Rates of both sexes (probability of dying between 15 and 60 years per 1000 population).
- infant.deaths (Numerical/Integer): Number of Infant Deaths per 1000 population.
- Alcohol (Numerical): Alcohol, recorded per capita (15+) consumption (in litres of pure alcohol).
- percentage.expenditure (Numerical): Expenditure on health as a percentage of Gross Domestic Product per capita(%).
- Hepatitis.B (Numerical/Integer): Hepatitis B (HepB) immunization coverage among 1-year-olds (%).
- Measles (Numerical/Integer): number of reported cases of measels per 1000 population.
- BMI (Numerical): Average Body Mass Index of entire population.
- under.five.deaths (Numerical/Integer): Number of under-five deaths per 1000 population.
- Polio (Numerical/Integer): Polio (Pol3) immunization coverage among 1-year-olds (%).
- Total.expenditure (Numerical): General government expenditure on health as a percentage of total government expenditure (%).
- Diphtheria (Numerical/Integer): Diphtheria tetanus toxoid and pertussis (DTP3) immunization coverage among 1-year-olds (%).
- HIV.AIDS (Numerical): Deaths per 1 000 live births HIV/AIDS (0-4 years).
- GDP (Numerical): Gross Domestic Product per capita (in USD).
- Population (Numerical): Population of the country.
- thinness..1.19.years (Numerical): Prevalence of thinness among children and adolescents for Age 10 to 19 (%).
- thinness.5.9.years (Numerical): Prevalence of thinness among children for Age 5 to 9(%).
- Income.composition.of.resources (Numerical): Human Development Index in terms of income composition of resources (index ranging from 0 to 1).
- Schooling (Numerical): Number of years of Schooling(years).
Identify the response variable (Y) and candidate predictors:
- Response Variable (Y): Life.expectancy
- Candidate Predictors (X_i): Status, Adult.Mortality, infant.deaths, Schooling, Population
Sample size (after split & removal of na values):
nrow(train_data)## [1] 1154- There are 1,649 observations after removing missing entries. We will be using the train_data as our sample size, which is 70% of the LE data and has 1,154 observations. The seed being used is 127.
Step 3: Descriptive Statistics
Numerical Variables
# Year Summary Stats
train_data %>%
summarise(
Year_mean = mean(Year, na.rm = TRUE),
Year_median = median(Year, na.rm = TRUE),
Year_sd = sd(Year, na.rm = TRUE),
Year_min = min(Year, na.rm = TRUE),
Year_Q1 = quantile(Year, 0.25, na.rm = TRUE),
Year_Q3 = quantile(Year, 0.75, na.rm = TRUE),
Year_max = max(Year, na.rm = TRUE)
)## Year_mean Year_median Year_sd Year_min Year_Q1 Year_Q3 Year_max
## 1 2007.879 2008 4.1054 2000 2005 2011 2015# Life Expectancy Summary Stats
train_data %>%
summarise(
Life.expectancy_mean = mean(Life.expectancy, na.rm = TRUE),
Life.expectancy_median = median(Life.expectancy, na.rm = TRUE),
Life.expectancy_sd = sd(Life.expectancy, na.rm = TRUE),
Life.expectancy_min = min(Life.expectancy, na.rm = TRUE),
Life.expectancy_Q1 = quantile(Life.expectancy, 0.25, na.rm = TRUE),
Life.expectancy_Q3 = quantile(Life.expectancy, 0.75, na.rm = TRUE),
Life.expectancy_max = max(Life.expectancy, na.rm = TRUE)
)## Life.expectancy_mean Life.expectancy_median Life.expectancy_sd
## 1 69.28735 71.7 8.708508
## Life.expectancy_min Life.expectancy_Q1 Life.expectancy_Q3 Life.expectancy_max
## 1 44 64.3 74.9 89# Adult Mortality Summary Stats
train_data %>%
summarise(
Adult.Mortality_mean = mean(Adult.Mortality, na.rm = TRUE),
Adult.Mortality_median = median(Adult.Mortality, na.rm = TRUE),
Adult.Mortality_sd = sd(Adult.Mortality, na.rm = TRUE),
Adult.Mortality_min = min(Adult.Mortality, na.rm = TRUE),
Adult.Mortality_Q1 = quantile(Adult.Mortality, 0.25, na.rm = TRUE),
Adult.Mortality_Q3 = quantile(Adult.Mortality, 0.75, na.rm = TRUE),
Adult.Mortality_max = max(Adult.Mortality, na.rm = TRUE)
)## Adult.Mortality_mean Adult.Mortality_median Adult.Mortality_sd
## 1 164.3068 148 120.8728
## Adult.Mortality_min Adult.Mortality_Q1 Adult.Mortality_Q3 Adult.Mortality_max
## 1 1 74 225 723# Infant Deaths Summary Stats
train_data %>%
summarise(
infant.deaths_mean = mean(infant.deaths, na.rm = TRUE),
infant.deaths_median = median(infant.deaths, na.rm = TRUE),
infant.deaths_sd = sd(infant.deaths, na.rm = TRUE),
infant.deaths_min = min(infant.deaths, na.rm = TRUE),
infant.deaths_Q1 = quantile(infant.deaths, 0.25, na.rm = TRUE),
infant.deaths_Q3 = quantile(infant.deaths, 0.75, na.rm = TRUE),
infant.deaths_max = max(infant.deaths, na.rm = TRUE)
)## infant.deaths_mean infant.deaths_median infant.deaths_sd infant.deaths_min
## 1 33.12132 3 124.7535 0
## infant.deaths_Q1 infant.deaths_Q3 infant.deaths_max
## 1 1 23 1600# Alcohol Summary Stats
train_data %>%
summarise(
Alcohol_mean = mean(Alcohol, na.rm = TRUE),
Alcohol_median = median(Alcohol, na.rm = TRUE),
Alcohol_sd = sd(Alcohol, na.rm = TRUE),
Alcohol_min = min(Alcohol, na.rm = TRUE),
Alcohol_Q1 = quantile(Alcohol, 0.25, na.rm = TRUE),
Alcohol_Q3 = quantile(Alcohol, 0.75, na.rm = TRUE),
Alcohol_max = max(Alcohol, na.rm = TRUE)
)## Alcohol_mean Alcohol_median Alcohol_sd Alcohol_min Alcohol_Q1 Alcohol_Q3
## 1 4.39753 3.69 3.942453 0.01 0.685 7.1375
## Alcohol_max
## 1 17.87# Percentage Ependiture Summary Stats
train_data %>%
summarise(
percentage.expenditure_mean = mean(percentage.expenditure, na.rm = TRUE),
percentage.expenditure_median = median(percentage.expenditure, na.rm = TRUE),
percentage.expenditure_sd = sd(percentage.expenditure, na.rm = TRUE),
percentage.expenditure_min = min(percentage.expenditure, na.rm = TRUE),
percentage.expenditure_Q1 = quantile(percentage.expenditure, 0.25, na.rm = TRUE),
percentage.expenditure_Q3 = quantile(percentage.expenditure, 0.75, na.rm = TRUE),
percentage.expenditure_max = max(percentage.expenditure, na.rm = TRUE)
)## percentage.expenditure_mean percentage.expenditure_median
## 1 651.3056 145.0437
## percentage.expenditure_sd percentage.expenditure_min
## 1 1629.372 0
## percentage.expenditure_Q1 percentage.expenditure_Q3
## 1 38.1045 501.721
## percentage.expenditure_max
## 1 18961.35# Hepatitis B Summary Stats
train_data %>%
summarise(
Hepatitis.B_mean = mean(Hepatitis.B, na.rm = TRUE),
Hepatitis.B_median = median(Hepatitis.B, na.rm = TRUE),
Hepatitis.B_sd = sd(Hepatitis.B, na.rm = TRUE),
Hepatitis.B_min = min(Hepatitis.B, na.rm = TRUE),
Hepatitis.B_Q1 = quantile(Hepatitis.B, 0.25, na.rm = TRUE),
Hepatitis.B_Q3 = quantile(Hepatitis.B, 0.75, na.rm = TRUE),
Hepatitis.B_max = max(Hepatitis.B, na.rm = TRUE)
)## Hepatitis.B_mean Hepatitis.B_median Hepatitis.B_sd Hepatitis.B_min
## 1 79.08232 89 25.74104 2
## Hepatitis.B_Q1 Hepatitis.B_Q3 Hepatitis.B_max
## 1 74 96 99# Measles Summary Stats
train_data %>%
summarise(
Measles_mean = mean(Measles, na.rm = TRUE),
Measles_median = median(Measles, na.rm = TRUE),
Measles_sd = sd(Measles, na.rm = TRUE),
Measles_min = min(Measles, na.rm = TRUE),
Measles_Q1 = quantile(Measles, 0.25, na.rm = TRUE),
Measles_Q3 = quantile(Measles, 0.75, na.rm = TRUE),
Measles_max = max(Measles, na.rm = TRUE)
)## Measles_mean Measles_median Measles_sd Measles_min Measles_Q1 Measles_Q3
## 1 2192.435 18 9995.222 0 0 394
## Measles_max
## 1 131441# BMI Summary Stats
train_data %>%
summarise(
BMI_mean = mean(BMI, na.rm = TRUE),
BMI_median = median(BMI, na.rm = TRUE),
BMI_sd = sd(BMI, na.rm = TRUE),
BMI_min = min(BMI, na.rm = TRUE),
BMI_Q1 = quantile(BMI, 0.25, na.rm = TRUE),
BMI_Q3 = quantile(BMI, 0.75, na.rm = TRUE),
BMI_max = max(BMI, na.rm = TRUE)
)## BMI_mean BMI_median BMI_sd BMI_min BMI_Q1 BMI_Q3 BMI_max
## 1 37.92998 43.55 19.82661 2.1 19.025 55.7 77.1# Under Five Deaths Summary Stats
train_data %>%
summarise(
under.five.deaths_mean = mean(under.five.deaths, na.rm = TRUE),
under.five.deaths_median = median(under.five.deaths, na.rm = TRUE),
under.five.deaths_sd = sd(under.five.deaths, na.rm = TRUE),
under.five.deaths_min = min(under.five.deaths, na.rm = TRUE),
under.five.deaths_Q1 = quantile(under.five.deaths, 0.25, na.rm = TRUE),
under.five.deaths_Q3 = quantile(under.five.deaths, 0.75, na.rm = TRUE),
under.five.deaths_max = max(under.five.deaths, na.rm = TRUE)
)## under.five.deaths_mean under.five.deaths_median under.five.deaths_sd
## 1 44.94627 4 167.3134
## under.five.deaths_min under.five.deaths_Q1 under.five.deaths_Q3
## 1 0 1 32
## under.five.deaths_max
## 1 2100# Polio Summary Stats
train_data %>%
summarise(
Polio_mean = mean(Polio, na.rm = TRUE),
Polio_median = median(Polio, na.rm = TRUE),
Polio_sd = sd(Polio, na.rm = TRUE),
Polio_min = min(Polio, na.rm = TRUE),
Polio_Q1 = quantile(Polio, 0.25, na.rm = TRUE),
Polio_Q3 = quantile(Polio, 0.75, na.rm = TRUE),
Polio_max = max(Polio, na.rm = TRUE)
)## Polio_mean Polio_median Polio_sd Polio_min Polio_Q1 Polio_Q3 Polio_max
## 1 82.79289 93 23.54729 3 79 97 99# Total Expenditure Summary Stats
train_data %>%
summarise(
Total.expenditure_mean = mean(Total.expenditure, na.rm = TRUE),
Total.expenditure_median = median(Total.expenditure, na.rm = TRUE),
Total.expenditure_sd = sd(Total.expenditure, na.rm = TRUE),
Total.expenditure_min = min(Total.expenditure, na.rm = TRUE),
Total.expenditure_Q1 = quantile(Total.expenditure, 0.25, na.rm = TRUE),
Total.expenditure_Q3 = quantile(Total.expenditure, 0.75, na.rm = TRUE),
Total.expenditure_max = max(Total.expenditure, na.rm = TRUE)
)## Total.expenditure_mean Total.expenditure_median Total.expenditure_sd
## 1 5.982045 5.89 2.305812
## Total.expenditure_min Total.expenditure_Q1 Total.expenditure_Q3
## 1 0.74 4.39 7.5225
## Total.expenditure_max
## 1 14.39# Diptheria Summary Stats
train_data %>%
summarise(
Diphtheria_mean = mean(Diphtheria, na.rm = TRUE),
Diphtheria_median = median(Diphtheria, na.rm = TRUE),
Diphtheria_sd = sd(Diphtheria, na.rm = TRUE),
Diphtheria_min = min(Diphtheria, na.rm = TRUE),
Diphtheria_Q1 = quantile(Diphtheria, 0.25, na.rm = TRUE),
Diphtheria_Q3 = quantile(Diphtheria, 0.75, na.rm = TRUE),
Diphtheria_max = max(Diphtheria, na.rm = TRUE)
)## Diphtheria_mean Diphtheria_median Diphtheria_sd Diphtheria_min Diphtheria_Q1
## 1 83.83016 92 21.90112 2 81
## Diphtheria_Q3 Diphtheria_max
## 1 97 99# HIV AIDS Summary Stats
train_data %>%
summarise(
HIV.AIDS_mean = mean(HIV.AIDS, na.rm = TRUE),
HIV.AIDS_median = median(HIV.AIDS, na.rm = TRUE),
HIV.AIDS_sd = sd(HIV.AIDS, na.rm = TRUE),
HIV.AIDS_min = min(HIV.AIDS, na.rm = TRUE),
HIV.AIDS_Q1 = quantile(HIV.AIDS, 0.25, na.rm = TRUE),
HIV.AIDS_Q3 = quantile(HIV.AIDS, 0.75, na.rm = TRUE),
HIV.AIDS_max = max(HIV.AIDS, na.rm = TRUE)
)## HIV.AIDS_mean HIV.AIDS_median HIV.AIDS_sd HIV.AIDS_min HIV.AIDS_Q1
## 1 1.880069 0.1 5.69773 0.1 0.1
## HIV.AIDS_Q3 HIV.AIDS_max
## 1 0.7 50.6# GDP Summary Stats
train_data %>%
summarise(
GDP_mean = mean(GDP, na.rm = TRUE),
GDP_median = median(GDP, na.rm = TRUE),
GDP_sd = sd(GDP, na.rm = TRUE),
GDP_min = min(GDP, na.rm = TRUE),
GDP_Q1 = quantile(GDP, 0.25, na.rm = TRUE),
GDP_Q3 = quantile(GDP, 0.75, na.rm = TRUE),
GDP_max = max(GDP, na.rm = TRUE)
)## GDP_mean GDP_median GDP_sd GDP_min GDP_Q1 GDP_Q3 GDP_max
## 1 5274.42 1522.676 10656.6 1.68135 455.3507 4674.074 114293.8# Population Summary Stats
train_data %>%
summarise(
Population_mean = mean(Population, na.rm = TRUE),
Population_median = median(Population, na.rm = TRUE),
Population_sd = sd(Population, na.rm = TRUE),
Population_min = min(Population, na.rm = TRUE),
Population_Q1 = quantile(Population, 0.25, na.rm = TRUE),
Population_Q3 = quantile(Population, 0.75, na.rm = TRUE),
Population_max = max(Population, na.rm = TRUE)
)## Population_mean Population_median Population_sd Population_min Population_Q1
## 1 14819584 1344918 72573412 36 182596
## Population_Q3 Population_max
## 1 7524793 1179681239# Thinness 10-19 Summary Stats
train_data %>%
summarise(
thinness..1.19.years_mean = mean(thinness..1.19.years, na.rm = TRUE),
thinness..1.19.years_median = median(thinness..1.19.years, na.rm = TRUE),
thinness..1.19.years_sd = sd(thinness..1.19.years, na.rm = TRUE),
thinness..1.19.years_min = min(thinness..1.19.years, na.rm = TRUE),
thinness..1.19.years_Q1 = quantile(thinness..1.19.years, 0.25, na.rm = TRUE),
thinness..1.19.years_Q3 = quantile(thinness..1.19.years, 0.75, na.rm = TRUE),
thinness..1.19.years_max = max(thinness..1.19.years, na.rm = TRUE)
)## thinness..1.19.years_mean thinness..1.19.years_median thinness..1.19.years_sd
## 1 4.863692 3 4.642256
## thinness..1.19.years_min thinness..1.19.years_Q1 thinness..1.19.years_Q3
## 1 0.1 1.6 7
## thinness..1.19.years_max
## 1 27.2# Thinness 5-9 Summary Stats
train_data %>%
summarise(
thinness.5.9.years_mean = mean(thinness.5.9.years, na.rm = TRUE),
thinness.5.9.years_median = median(thinness.5.9.years, na.rm = TRUE),
thinness.5.9.years_sd = sd(thinness.5.9.years, na.rm = TRUE),
thinness.5.9.years_min = min(thinness.5.9.years, na.rm = TRUE),
thinness.5.9.years_Q1 = quantile(thinness.5.9.years, 0.25, na.rm = TRUE),
thinness.5.9.years_Q3 = quantile(thinness.5.9.years, 0.75, na.rm = TRUE),
thinness.5.9.years_max = max(thinness.5.9.years, na.rm = TRUE)
)## thinness.5.9.years_mean thinness.5.9.years_median thinness.5.9.years_sd
## 1 4.941334 3.1 4.716583
## thinness.5.9.years_min thinness.5.9.years_Q1 thinness.5.9.years_Q3
## 1 0.1 1.6 7.1
## thinness.5.9.years_max
## 1 28.2# Income Composition of Resources Summary Stats
train_data %>%
summarise(
Income.composition.of.resources_mean = mean(Income.composition.of.resources, na.rm = TRUE),
Income.composition.of.resources_median = median(Income.composition.of.resources, na.rm = TRUE),
Income.composition.of.resources_sd = sd(Income.composition.of.resources, na.rm = TRUE),
Income.composition.of.resources_min = min(Income.composition.of.resources, na.rm = TRUE),
Income.composition.of.resources_Q1 = quantile(Income.composition.of.resources, 0.25, na.rm = TRUE),
Income.composition.of.resources_Q3 = quantile(Income.composition.of.resources, 0.75, na.rm = TRUE),
Income.composition.of.resources_max = max(Income.composition.of.resources, na.rm = TRUE)
)## Income.composition.of.resources_mean Income.composition.of.resources_median
## 1 0.6311144 0.673
## Income.composition.of.resources_sd Income.composition.of.resources_min
## 1 0.1803673 0
## Income.composition.of.resources_Q1 Income.composition.of.resources_Q3
## 1 0.50625 0.748
## Income.composition.of.resources_max
## 1 0.936# Schooling Summary Stats
train_data %>%
summarise(
Schooling_mean = mean(Schooling, na.rm = TRUE),
Schooling_median = median(Schooling, na.rm = TRUE),
Schooling_sd = sd(Schooling, na.rm = TRUE),
Schooling_min = min(Schooling, na.rm = TRUE),
Schooling_Q1 = quantile(Schooling, 0.25, na.rm = TRUE),
Schooling_Q3 = quantile(Schooling, 0.75, na.rm = TRUE),
Schooling_max = max(Schooling, na.rm = TRUE)
)## Schooling_mean Schooling_median Schooling_sd Schooling_min Schooling_Q1
## 1 12.0714 12.2 2.817797 4.4 10.4
## Schooling_Q3 Schooling_max
## 1 13.875 20.7Numerical Variable Interpretations:
- Year – median year 2008; no skew in data.
- Life expectancy – median ~ 71.7 years & mean 69.28; normally distributed.
- Adult Mortality – sd ~ 120.87 deaths/1000; large cross-country differences.
- Infant deaths – median ~ 3, but max 1600; strongly right-skewed.
- Alcohol – sd ~ 3.94; wide variation across countries.
- Percentage expenditure – median 145, but max 18,961; strongly right-skewed.
- Hepatitis B – median 89% vaccinated; generally high but inconsistent.
- Measles – median 18 cases, but outbreaks raise max >130,000.
- BMI – median 43.6; reflects differences in nutrition and development.
- Under-five deaths – median 4, but can exceed 2000; strongly right-skewed.
- Polio – median 93% coverage; high immunization overall.
- Total expenditure – median 5.9% of government spending; moderate variation.
- Diphtheria – median 92% coverage; parallels Polio and Hep B patterns.
- HIV/AIDS deaths – median 0.1 but max 50.6; right-skewed burden.
- GDP – median $1,523, mean ~ $5,274; indicates strong right-skew and inequality.
- Population – median 1.34 million, max >1 billion; extremely skewed.
- Thinness (1–19 yrs) – median 3%; higher rates in lower-income regions.
- Thinness (5–9 yrs) – median 3.1%; similar pattern to age 1–19 group.
- Income composition of resources – median 0.67 on a 0–1 scale; reflects developmental variation.
- Schooling – median 12.2 years; indicates differing education access globally.
Categorical Variables
# Country
## Frequency Table
country_freq <- table(train_data$Country)
country_freq##
## Afghanistan Albania Algeria
## 15 15 9
## Angola Argentina Armenia
## 5 11 11
## Australia Austria Azerbaijan
## 10 7 10
## Bangladesh Belarus Belgium
## 9 11 9
## Belize Benin Bhutan
## 12 9 8
## Bosnia and Herzegovina Botswana Brazil
## 7 9 9
## Bulgaria Burkina Faso Burundi
## 12 9 9
## Cabo Verde Cambodia Cameroon
## 6 7 7
## Canada Central African Republic Chad
## 11 3 6
## Chile China Colombia
## 6 9 11
## Comoros Costa Rica Croatia
## 9 11 7
## Cyprus Djibouti Dominican Republic
## 11 5 7
## Ecuador El Salvador Equatorial Guinea
## 11 8 1
## Eritrea Estonia Ethiopia
## 9 6 6
## Fiji France Gabon
## 11 10 6
## Georgia Germany Ghana
## 13 8 10
## Greece Guatemala Guinea
## 11 7 5
## Guinea-Bissau Guyana Haiti
## 3 9 1
## Honduras India Indonesia
## 7 8 10
## Iraq Ireland Israel
## 10 3 12
## Italy Jamaica Jordan
## 11 9 13
## Kazakhstan Kenya Kiribati
## 12 12 11
## Latvia Lebanon Lesotho
## 10 12 8
## Liberia Lithuania Luxembourg
## 4 10 8
## Madagascar Malawi Malaysia
## 6 11 11
## Maldives Mali Malta
## 9 9 8
## Mauritania Mauritius Mexico
## 8 13 13
## Mongolia Montenegro Morocco
## 11 6 9
## Mozambique Myanmar Namibia
## 9 10 3
## Nepal Netherlands Nicaragua
## 10 2 13
## Niger Nigeria Pakistan
## 2 6 9
## Panama Papua New Guinea Paraguay
## 12 13 9
## Peru Philippines Poland
## 7 11 9
## Portugal Romania Russian Federation
## 9 10 5
## Rwanda Samoa Sao Tome and Principe
## 11 10 6
## Senegal Serbia Seychelles
## 6 9 9
## Sierra Leone Solomon Islands South Africa
## 4 7 12
## Spain Sri Lanka Suriname
## 8 8 9
## Swaziland Sweden Syrian Arab Republic
## 11 3 6
## Tajikistan Thailand Timor-Leste
## 6 10 5
## Togo Tonga Trinidad and Tobago
## 5 12 8
## Tunisia Turkey Turkmenistan
## 10 6 10
## Uganda Ukraine Uruguay
## 10 12 12
## Uzbekistan Vanuatu Zambia
## 11 13 6
## Zimbabwe
## 9## Percentage Table
round(prop.table(country_freq) * 100, 2)##
## Afghanistan Albania Algeria
## 1.30 1.30 0.78
## Angola Argentina Armenia
## 0.43 0.95 0.95
## Australia Austria Azerbaijan
## 0.87 0.61 0.87
## Bangladesh Belarus Belgium
## 0.78 0.95 0.78
## Belize Benin Bhutan
## 1.04 0.78 0.69
## Bosnia and Herzegovina Botswana Brazil
## 0.61 0.78 0.78
## Bulgaria Burkina Faso Burundi
## 1.04 0.78 0.78
## Cabo Verde Cambodia Cameroon
## 0.52 0.61 0.61
## Canada Central African Republic Chad
## 0.95 0.26 0.52
## Chile China Colombia
## 0.52 0.78 0.95
## Comoros Costa Rica Croatia
## 0.78 0.95 0.61
## Cyprus Djibouti Dominican Republic
## 0.95 0.43 0.61
## Ecuador El Salvador Equatorial Guinea
## 0.95 0.69 0.09
## Eritrea Estonia Ethiopia
## 0.78 0.52 0.52
## Fiji France Gabon
## 0.95 0.87 0.52
## Georgia Germany Ghana
## 1.13 0.69 0.87
## Greece Guatemala Guinea
## 0.95 0.61 0.43
## Guinea-Bissau Guyana Haiti
## 0.26 0.78 0.09
## Honduras India Indonesia
## 0.61 0.69 0.87
## Iraq Ireland Israel
## 0.87 0.26 1.04
## Italy Jamaica Jordan
## 0.95 0.78 1.13
## Kazakhstan Kenya Kiribati
## 1.04 1.04 0.95
## Latvia Lebanon Lesotho
## 0.87 1.04 0.69
## Liberia Lithuania Luxembourg
## 0.35 0.87 0.69
## Madagascar Malawi Malaysia
## 0.52 0.95 0.95
## Maldives Mali Malta
## 0.78 0.78 0.69
## Mauritania Mauritius Mexico
## 0.69 1.13 1.13
## Mongolia Montenegro Morocco
## 0.95 0.52 0.78
## Mozambique Myanmar Namibia
## 0.78 0.87 0.26
## Nepal Netherlands Nicaragua
## 0.87 0.17 1.13
## Niger Nigeria Pakistan
## 0.17 0.52 0.78
## Panama Papua New Guinea Paraguay
## 1.04 1.13 0.78
## Peru Philippines Poland
## 0.61 0.95 0.78
## Portugal Romania Russian Federation
## 0.78 0.87 0.43
## Rwanda Samoa Sao Tome and Principe
## 0.95 0.87 0.52
## Senegal Serbia Seychelles
## 0.52 0.78 0.78
## Sierra Leone Solomon Islands South Africa
## 0.35 0.61 1.04
## Spain Sri Lanka Suriname
## 0.69 0.69 0.78
## Swaziland Sweden Syrian Arab Republic
## 0.95 0.26 0.52
## Tajikistan Thailand Timor-Leste
## 0.52 0.87 0.43
## Togo Tonga Trinidad and Tobago
## 0.43 1.04 0.69
## Tunisia Turkey Turkmenistan
## 0.87 0.52 0.87
## Uganda Ukraine Uruguay
## 0.87 1.04 1.04
## Uzbekistan Vanuatu Zambia
## 0.95 1.13 0.52
## Zimbabwe
## 0.78# Status
## Frequency Table
status_freq <- table(train_data$Status)
status_freq##
## Developed Developing
## 155 999## Percentage Table
round(prop.table(status_freq) * 100, 2)##
## Developed Developing
## 13.43 86.57Categorical Variable Interpretations:
- Country - The frequency of a certain country ranges from 6-15, the frequency 15 indicating all the years from 2000-2015 of a country’s observations are included. Any less than 15 means certain observations were removed during data cleaning to get rid of the present NA values.
- Status - Developing countries make up about 86.6% of the training data, while Developed countries make up about 13.4%. This means the dataset is weighted toward lower-income nations, which may affect comparisons of health indicators and life expectancy.
Step 4: Visual Exploration
Numerical Variable Histograms
par(mfrow = c(4, 5), mar = c(2, 2, 2, 1))
hist(train_data$Year, main = "Year")
hist(train_data$Life.expectancy, main = "Life Expectancy")
hist(train_data$Adult.Mortality, main = "Adult Mortality")
hist(train_data$infant.deaths, main = "Infant Deaths")
hist(train_data$Alcohol, main = "Alcohol")
hist(train_data$percentage.expenditure, main = "Percentage Expenditure")
hist(train_data$Hepatitis.B, main = "Hepatitis B")
hist(train_data$Measles, main = "Measles")
hist(train_data$BMI, main = "BMI")
hist(train_data$under.five.deaths, main = "Under-Five Deaths")
hist(train_data$Polio, main = "Polio")
hist(train_data$Total.expenditure, main = "Total Expenditure")
hist(train_data$Diphtheria, main = "Diphtheria")
hist(train_data$HIV.AIDS, main = "HIV/AIDS")
hist(train_data$GDP, main = "GDP")
hist(train_data$Population, main = "Population")
hist(train_data$thinness..1.19.years, main = "Thinness (10–19)")
hist(train_data$thinness.5.9.years, main = "Thinness (5–9)")
hist(train_data$Income.composition.of.resources, main = "Income Composition")
hist(train_data$Schooling, main = "Schooling")
Categorical Variable Bar Charts
barplot(table(train_data$Status), main = "Status", col = c("blue", "red"))
barplot(table(train_data$Country), main = "Country")
vars <- c("Life.expectancy", "Adult.Mortality", "Schooling", "infant.deaths", "Population")
cor_matrix <- cor(train_data[, vars], use = "complete.obs")
corrplot(cor_matrix,
method = "color",
type = "lower",
addCoef.col = "black",
number.cex = 0.7,
tl.col = "black",
tl.srt = 45,
diag = FALSE,
title = "Correlation Matrix (Training Data)",
mar = c(0, 0, 2, 0))