Data Analysis Report of Health and Population Statistics of East Asian and Pacific Countries

This report analyses the health and population statistics of East Asian and Pacific countries from 2001 to 2015. The report includes one-variable and two-variable analysis, k-means clustering analysis, and linear regression analysis. The report also discusses the limitations and scope of the study. The outcome of the analysis can be used by planners, governments to assess the relation of health and various attributes of health of the region.

Data analysis report of the health and population statistics of East Asian and Pacific countries
Name of the Student

Table of Contents
1 Introduction..................................................................................................................................1
1.1 Authorisation and Purpose...............................................................................................1
Limitations...................................................................................................................................1
Scope............................................................................................................................................1
Methodology...............................................................................................................................1
2 Data Setup....................................................................................................................................1
3 Exploratory Data Analysis.............................................................................................................2
3.1 One Variable Analysis................................................................................................................2
3.1.1 One Variable Analysis – 1.......................................................................................................2
3.1.2 One Variable Analysis – 2.......................................................................................................3
3.1.3 One Variable Analysis – 3.......................................................................................................5
3.2 Two-variable analysis.................................................................................................................6
3.2.1 Two-variable analysis 1...........................................................................................................6
3.2.2 Two-variable analysis 2...........................................................................................................7
4 Advanced analysis.........................................................................................................................9
4.1 Clustering...................................................................................................................................9
4.1.1 Brief explanation of k-means and clustering..........................................................................9
4.1.2 Clustering Analysis..................................................................................................................9
4.2 Linear regression......................................................................................................................11
4.2.1 Brief definition of linear regression......................................................................................11
4.2.2 Linear Regression 1...............................................................................................................11
4.2.3 Linear Regression 2...............................................................................................................13
5 Conclusion...................................................................................................................................14
6 Reflection....................................................................................................................................14
Reference.......................................................................................................................................16
Page | ii

1 Introduction
1.1 Authorisation and Purpose
The endeavour of the present study is to investigate the health of East Asia and Pacific region.
Information for the period of 2001 to 2015 is taken for the analysis. For the purpose of the
study data has been collected from World Bank. The outcome of the present analysis can be
used by planners, governments to assess the relation of health and various attributes of health
of the region.
Limitations
The facts presented for the analysis pertains to the region of East Asia and Pacific only. Further,
the data pertains to the time period of 2001 to 2015. In addition, information on certain
attributes are missing for some countries.
Scope
The facts for the present analysis consists of 26 attributes concerning health of the region. The
information pertains to East Asia and Pacific region for the period of 2001 to 2015.
The scope of the present study is limited by the presence of missing data. Statistical analysis of
the data is undertaken. Graphs are used to provide a visual representation of the statistical
analysis.
Four different types of graphs are used for the analysis. One-variable and two-variable analysis
are used to represent the attributes. k-means clustering analysis is used to relate the countries
of the region. Linear regression analysis is used to find the relation between attributes.
Methodology
Data collected from World Bank on the health aspects of the East Asia and Pacific region for the
period of 2001 to 2015 is used for the analysis.
2 Data Setup
Prior to starting of the analysis it is important that the data file be imported in “R-program.” In
addition, the “R-program” would require library files for the analysis. Thus the first lines of code
is used for importing the data file into “R.”
The first line also provides information to the “R-program” –
1. The data file to be imported is a CSV file
2. The first row of information is to be taken as header
3. The values are separated by comma
4. There are missing values

3 Exploratory Data Analysis
3.1 One Variable Analysis
3.1.1 One Variable Analysis – 1
The distribution of total unemployment of East Asia and Pacific Region from 2001 to 2015 is
first analysed. From statistical analysis that the average unemployment in the last 14 years is
4.29% with a minimum of 0.1% to a maximum of 11.9%. The standard deviation of
unemployment is 2.22%. Boxplot is used to study the distribution of unemployment. From the
boxplot it is found that unemployment is skewed right. Moreover, the rate of unemployment in
some of the countries is very high.
Page | 2
Analysis1 <- filter(Health, Series.Code %in% "SL.UEM.TOTL.ZS")
describe(Analysis1$value)
jpeg("Analysis1.jpeg")
Analysis1 <- ggplot(Analysis1, aes(x = factor(0), y = value)) + geom_boxplot()
Analysis1 <- Analysis1 + xlab("Unemployment, total (% of total labor force)") +
scale_x_discrete(breaks = NULL) + coord_flip()
Analysis1 <- Analysis1 + ggtitle("Distribution of Unemployment, \n total (% of total labor
force) from 2001 to 2014") + theme_bw()
print(Analysis1)
dev.off()
# Loading the required data
Health <- read.csv(file.choose(), header = TRUE, sep = "," , na.strings = "..", blank.lines.skip =
TRUE,)
# Loading required library files
library(dplyr)
library(ggplot2)
library(psych)
library(reshape2)
library(cluster)

3.1.2 One Variable Analysis – 2
The distribution of GNI per capita of East Asia and Pacific Region from 2001 to 2015 is next
analysed. From statistical analysis that the average GNI per capita in the last 14 years is
$11522.45 with standard deviation of $15406. The minimum and maximum GNI per capita are
$310 and $76300 respectively. Boxplot is used to study the distribution of GNI per capita for the
last 14 years. From the boxplot it is found that GIN is skewed right. Moreover, the GNI per
capita in some of the countries is very high.
Page | 3

Page | 4
Analysis2 <- filter(Health, Series.Code %in% "NY.GNP.PCAP.CD")
describe(Analysis2$value)
jpeg("Analysis2.jpeg")
Analysis2 <- ggplot(Analysis2, aes(x = factor(0), y = value)) + geom_boxplot()
Analysis2 <- Analysis2 + xlab("GNI per capita") + scale_x_discrete(breaks = NULL) +
coord_flip()
Analysis2 <- Analysis2 + ggtitle("Distribution of GNI per capita from 2001 to 2014") +
theme_bw()
print(Analysis2)

3.1.3 One Variable Analysis – 3
The total unemployment of the region (as % of labour force) is analysed through a Histogram.
From the histogram it is found that the unemployment is left skewed. Thus the average
unemployment is higher than the median unemployment rate.
Page | 5
j jpeg("Analysis3.jpeg")
fill <- "Orange"
barlines <- "blue"
Health <- na.omit(Health)
Analysis3 <- ggplot(Health, aes(x = value)) + geom_histogram(aes(y = ..count..), binwidth =
1.5,colour = barlines, fill = fill)
Analysis3 <- Analysis3 + scale_x_continuous(name = "Mean Unemployment Rate", breaks =
seq(0,12,1), limits = c(0,12) ) + scale_y_continuous(name = "Frequency")
Analysis3 <- Analysis3 + ggtitle("Distribution of Unemployment, \n total (% of total labor force)
from 2001 to 2014")
Analysis3 <- Analysis3 +theme(plot.title = element_text(hjust = 0.5)) + theme_bw()
print(Analysis3)

3.2 Two-variable analysis
3.2.1 Two-variable analysis 1
Boxplots are used for the study of two-variable analysis. The total unemployment (% of labour
force) of the East Asia and Pacific region for the period of 2001 to 2014 is used as a two-variable
analysis. The analysis of the boxplot shows that there is wide variation in unemployment
amongst countries of the region. moreover, the rate of unemployment for some countries has
wide variations.
Page | 6
Analysis4 <- filter(Health, Series.Code %in% "SL.UEM.TOTL.ZS")
jpeg("Analysis4.jpeg")
fill <- "Orange"
Analysis4 <- ggplot(Analysis4, aes(x = Country.Code, y = value)) + geom_boxplot(fill = fill)
Analysis4 <- Analysis4 + scale_x_discrete(name = "Country") + scale_y_continuous(name =
"Total Unemployment Rate")+ theme_bw()
Analysis4 <- Analysis4 +theme(axis.text.x = element_text(angle = 90, hjust = 1)) +
ggtitle("Distribution of Total Unemployment rate from 2001 to 2014 \n for countries of East
Asia and Pacific Region")
print(Analysis4)

3.2.2 Two-variable analysis 2
Boxplot is used to study the relation GNI per capita of the countries for the period from 2001 to
2014. The use of Boxplot offers the advantage that individual countries of the region can be
compared on the basis of the distribuion of their GNI for the 14 years of study. From the
Boxplot it is found that for most of the countries of the region GNI is less than $10000. For
countries with GNI above $10000 there are wide variations in values for the last 14 years.
Page | 7

Page | 8
Analysis5 <- filter(Health, Series.Code %in% "NY.GNP.PCAP.CD")
jpeg("Analysis5.jpeg")
fill <- "Orange"
Analysis5 <- ggplot(Analysis5, aes(x = Country.Code, y = value)) + geom_boxplot(fill = fill)
Analysis5 <- Analysis5 + scale_x_discrete(name = "Country") + scale_y_continuous(name = "Gross
National Income")+ theme_bw()
Analysis5 <- Analysis5 +theme(axis.text.x = element_text(angle = 90, hjust = 1)) +
ggtitle("Distribution of GNI per Capita from 2001 to 2014 for countries of East Asia and Pacific
Region")
print(Analysis5)

4 Advanced analysis
4.1 Clustering
4.1.1 Brief explanation of k-means and clustering
The method of segregating a data into analogous groups is known as clustering. The behaviour
of analogous groups is similar. Clustering involves dividing the data set into groups and thus
allocating data points to such groups (Oleiwi 2016). Data points in a group have similar
properties.
In k-means clustering the data values are segregated into k-groups. In k-means clustering first k-
centroids are identified. Each data is sequentially added to a centroid nearest to its value. When
a second data point is added the centroid is adjusted based on the mean of the two points. This
process is continued until all data points are exhausted (Witten et al., 2016).
4.1.2 Clustering Analysis
Page | 9
Health1 <- filter(Health, Series.Code %in% c("NY.GNP.PCAP.CD", "SL.UEM.TOTL.ZS"))
Health1 <- melt(Health1, Series.Code = c("Series.Code","Country.Name","Country.Code"))
Health1 <- dcast(Health1, formula = Country.Code ~ Series.Code, mean)
Health1 <- na.omit(Health1)
View(Health1)
grpdata <- kmeans(Health1[,c("NY.GNP.PCAP.CD", "SL.UEM.TOTL.ZS")],centers = 3, nstart = 10)
grpdata
o = order(grpdata$cluster)
data.frame(Health1$Country.Code[o], grpdata$cluster[o])
jpeg("Analysis6.jpeg")
Analysis6<- plot(Health1$NY.GNP.PCAP.CD, Health1$SL.UEM.TOTL.ZS, type="n",
xlim=c(0,50000), xlab="Gross National Income", ylab="Unemployment Rate")
text(x=Health1$NY.GNP.PCAP.CD, y=Health1$SL.UEM.TOTL.ZS,
labels=Health1$Country.Code,col=grpdata$cluster+1)
print(Analysis6)

For k-means clustering the GNI per capita is related to total unemployment for 2014.
The chart shows that the countries can be grouped into three groups.
Countries with very low GNI
Countries with average GNI and average Unemployment
Countries with High GNI and average Unemployment
Page | 10

4.2 Linear regression
4.2.1 Brief definition of linear regression
Simple Linear regression assumes that there is a relationship between two variables which can
be represented as a straight line. The two variables are denoted as response and predictor
variable (Braun et al., 2014). The predictor variable is used to predict the value of the response
variable. The relationship is represented as:
Y = β0 +β1 X
where
Y = Response Variable
X = Predictor Variable
β0 = Intercept
β1 = Slope
The line represented by the above formula predicts the relationship between the variables
(Asumadu-Sarkodie and Owusu 2017). Linear regression is used to predict the change in value
of “Y” for each unit change in “X.”
4.2.2 Linear Regression 1
In the first linear regression an attempt is made to predict the health expenditure per capita
from the gross national income per capita. The gross national income is the independent
variable and health expenditure is the dependent variable. The data from 2001 to 2014 is taken
into consideration. It can be safely assumed that with increase in incomes of the citizens of the
country there would be an increase in health expenditure of the country. The higher the health
expenditure the better would be the residents of the region. From the analysis it is found that
the health expenditure can be predicted as:
Health Expenditure = 0.075*GNI per capita – 31.48
From the above equation it is seen that with increase in GNI per capita of the citizens of the
countries there would be a corresponding increase in health expenditure. Further, 79.43% of
the variations in Health Expenditure of the countries can be predicted from GNI per capita.
Page | 11

Page | 12
Analysis7 <- lm(formula = SH.XPD.PCAP ~ NY.GNP.PCAP.CD , data = Health2)
summary(Analysis7)
jpeg("Analysis7.jpeg")
Analysis7 <- ggplot(Health2, aes(x = NY.GNP.PCAP.CD, y = SH.XPD.PCAP)) +
geom_point(shape = 4)
Analysis7 <- Analysis7 + scale_x_continuous(name = "GNI per Capita") +
scale_y_continuous(name = "Health Expenditure per Capita")+ geom_smooth(method=lm,
se = FALSE)
Analysis7 <- Analysis7 + theme_bw()+ ggtitle("Relation of Health Expenditure per Capita to
GNI per Capita")
Analysis7
print(Analysis7)

4.2.3 Linear Regression 2
In the second linear regression an attempt is made to predict the GNI per capita from log of
unemployment. The total unemployment (log) is the independent variable and GNI is the
dependent variable. The data from 2001 to 2014 is taken into consideration. From the analysis
it is found that the GNI per capita can be predicted as:
GNI per capita = 8.1927 + 0.1120*log(total unemployment)
From the above equation it is seen that with increase in Unemployment of the citizens of the
countries there would be a corresponding increase in GNI per capita. Further, 2.38% of the
variations in GNI per capita of the countries can be predicted from Unemployment.
Page | 13
Analysis8 <- lm(formula = log(NY.GNP.PCAP.CD) ~ SL.UEM.TOTL.ZS, data = Health1)
summary(Analysis8)
jpeg("Analysis8.jpeg")
Analysis8 <- ggplot(Health1, aes(x = SL.UEM.TOTL.ZS, y = log(NY.GNP.PCAP.CD))) +
geom_point(shape = 4)
Analysis8 <- Analysis8 + scale_x_continuous(name = "Unemployment Rate") +
scale_y_continuous(name = "Log of Gross National Income")+ geom_smooth(method=lm,
se = FALSE)
Analysis8 <- Analysis8 + theme_bw()+ ggtitle("Relation of Log of Gross National Income to
Unemployment Rate")
print(Analysis8)

5 Conclusion
The data provided by the world bank for the region of East Asia and Pacific region provided is
with important insights into the Health of the Area.
The analysis depicted that the total unemployment and GNI per capita is left skewed in the
region. Further, in some of the countries the rate of unemployment is very high. moreover, the
GNI for most of the countries of the region has been lower than $10000 in the last 14 years. In
addition, there are wide variations in total unemployment amongst the countries.
Further from the study it is found that with increase in GNI per capita there has been an
increase in health expenditure amongst the countries of the region.
6 Reflection
From the study of the health statistics of East Asia and Pacific region it is found that there is
wide variation in GNI and unemployment of the region. In addition, we could find that the
countries can be grouped into three groups based on their GNI and total unemployment. The
study of the present information was compounded by the presentation of the data. The
attributes were presented in a series requiring the need to reshape the data.
Page | 14

Reference
Asumadu-Sarkodie, S. and Owusu, P.A., 2017. Recent evidence of the relationship between
carbon dioxide emissions, energy use, GDP, and population in Ghana: A linear regression
approach. Energy Sources, Part B: Economics, Planning, and Policy, 12(6), pp.495-503.
Braun, M.R., Altan, H. and Beck, S.B.M., 2014. Using regression analysis to predict the future
energy consumption of a supermarket in the UK. Applied Energy, 130, pp.305-313.
Oleiwi, W.K., 2016. Using the Fuzzy Logic to Find Optimal Centers of Clusters of K-means.
International Journal of Electrical and Computer Engineering, 6(6), p.3068.
Witten, I.H., Frank, E., Hall, M.A. and Pal, C.J., 2016. Data Mining: Practical machine learning
tools and techniques. Morgan Kaufmann.
Page | 15