Project Final Report
Minh Tri Bui
17/05/2022
Import Your Data
In the following code hunk, importing and wrangling the data.
# Attach 3 datasets to R
setwd("D:/Statistics/R/R data/test")
malaria <- read_csv("R captone/Number of indigenous malaria cases.csv")
falciparum <- read_csv("R captone/Number of indigenous P. falciparum malaria cases.csv")
vivax <- read_csv("R captone/Number of indigenous P. vivax malaria cases.csv")
# Select the value that we are going to plot
malaria <- malaria %>% select(Location, Period, FactValueNumeric)
falciparum <- falciparum %>% select(Location, Period, FactValueNumeric)
vivax <- vivax %>% select(Location, Period, FactValueNumeric)
## Changing column names
colnames(malaria)[which(names(malaria) == "FactValueNumeric")] <- "Total_cases"
colnames(falciparum)[which(names(falciparum) == "FactValueNumeric")] <- "P.falciparum"
colnames(vivax)[which(names(vivax) == "FactValueNumeric")] <- "P.vivax"
## Merging dataframe
data <- malaria %>% full_join(falciparum) %>% full_join(vivax)
data$Others = data$Total_cases - data$P.falciparum - data$P.vivax
data$Period = as.factor(data$Period)
#Changing names of the specific attributes within Location:
data$Location[data$Location=="Democratic Republic of the Congo"] <- "Congo"
data$Location[data$Location=="United Republic of Tanzania"] <- "Tanzania"
data$Location[data$Location=="Venezuela (Bolivarian Republic of)"] <- "Venezuela"
#Show the table of the dataset
datatable(data,
filter = 'top',
caption = htmltools::tags$caption(
style = 'caption-side: top; text-align: center;',
'Table 1.1: ', htmltools::em('The number of indigenous malaria cases in the world')))Part 1: Top ten countries in 2020
1.1 The total malaria cases
For my first figure, I am going to create a bar chart that plots top ten countries with highest total malaria cases on the x axis (the cases total) and the number of malaria cases the y axis in 2020.
# Select top 10 countries for total cases
plot1 <- data %>% arrange(desc(Period),desc( Total_cases)) %>% slice(1:10)
datatable(plot1,
filter = 'top',
caption = htmltools::tags$caption(
style = 'caption-side: top; text-align: center;',
'Table 1.2: ', htmltools::em('The number of top ten malaria cases in the world')))# Format the e display for some columns
plot1$emalaria <- formatC(plot1$Total_cases, format = "e", digits = 2) # Note: class of "emalaria" is character, not numeric
plot1$efalciparum <- formatC(plot1$P.falciparum, format = "e", digits = 2)
plot1$evivax <- formatC(plot1$P.vivax, format = "e", digits = 2)
plot1$eothers <- formatC(plot1$Others, format = "e", digits = 2)
# Draw a plot
ggplot(plot1, aes(x = reorder(Location, - Total_cases), y = Total_cases, fill = -Total_cases)) +
geom_bar(stat = "identity") +
labs( x = "Country", y= "Number of cases", title = "Top ten countries with highest malaria cases in 2020") +
scale_y_continuous(limits = c(0, 2.5e+07), labels = label_scientific()) +
geom_text(aes(label = emalaria), vjust = - 0.5, size = 3) +
theme(plot.title = element_text(hjust = 0.5)) + # Adjust the title to the center
theme(plot.title = element_text(size = (15), face = "bold" ),
axis.title = element_text(size = (13), face = "bold", colour = "black"),
axis.text.x = element_text(angle= 30, hjust=1, size= 9))
For further examination, I would like to break down the total number of cases to cases caused by P.falcuparum, P.viax and other factors.
plot2 <-
plot1 %>% pivot_longer(cols = c("Total_cases","P.falciparum","P.vivax","Others"),
names_to = "Cases",
values_to = "Count",
values_drop_na = TRUE) %>%
select(- (emalaria :eothers))
plot2$ecount <- formatC(plot2$Count, format = "e", digits = 2)
plot2$Cases <- factor(plot2$Cases, levels = c ("Total_cases", "P.falciparum", "P.vivax", "Others"))
a <- ggplot(plot2, aes(x = reorder(Location, - Count), y = Count, text1 = ecount, fill = Cases)) +
geom_bar(stat = "identity", position = "dodge") +
labs( x = "Country", y= "Number of cases", title = "Top ten countries with malaria cases in 2020") +
scale_y_continuous(limits = c(0, 2.5e+07), labels = label_scientific()) +
theme(plot.title = element_text(hjust = 0.5)) + # Adjust the title to the center
theme(plot.title = element_text(size = (15), face = "bold" ),
axis.title = element_text(size = (13), face = "bold", colour = "black"),
axis.text.x = element_text(angle= 30, hjust=1, size= 9))
#Saving ggplotly:
# saveWidget(as_widget(ggplotly(a)), file="Top ten countries with malaria cases in 2020.html")
#Save on online system
# Sys.setenv("plotly_username"="Minh_Tri_92")
# Sys.setenv("plotly_api_key"="B6mso8qyGEpF7kJekgKe")
# api_create(ggplotly(a), "Top ten countries with malaria cases in 2020")
ggplotly (a)Result:
As we can see from table 1.2 or from interactive graph above, there is a lack data leading to the causes for malaria are unknown in 9/10 top ten countries.
Only Tanzania has full data with 5,678,149 P.falciparum cases and 291,804 other causes in 2020.
1.2 The P.falciparum cases
For my next figure, I am going to create a bar chart that plots top ten countries with highest malaria cases caused by P.falciparum on the x axis and the number of malaria cases the y axis in 2020.
# Select the top ten countries of P.falciparum cases
plot3 <- data %>% arrange(desc(Period),desc(P.falciparum)) %>% slice(1:10)
datatable(plot3,
filter = 'top',
caption = htmltools::tags$caption(
style = 'caption-side: top; text-align: center;',
'Table 1.3: ', htmltools::em('The number of top ten malaria cases caused by *P.falciparum*')))# Display e value for some columns
plot3$emalaria <- formatC(plot3$Total_cases, format = "e", digits = 2) # Note: class of "emalaria" is character, not numeric
plot3$efalciparum <- formatC(plot3$P.falciparum, format = "e", digits = 2)
plot3$evivax <- formatC(plot3$P.vivax, format = "e", digits = 2)
plot3$eothers <- formatC(plot3$Others, format = "e", digits = 2)
# Draw a plot
ggplot(plot3, aes(x = reorder(Location, - P.falciparum), y = P.falciparum, fill= -P.falciparum)) +
geom_bar(stat = "identity", color='blue') +
labs( x = "Country", y= "Number of cases", title = "Top ten countries with highest P.falciparum cases in 2020") +
geom_text(aes(label = efalciparum), vjust = - 0.5, size = 3) +
theme(plot.title = element_text(hjust = 0.5)) + # Adjust the title to the center
theme(plot.title = element_text(size = (15), face = "bold" ),
axis.title = element_text(size = (13), face = "bold", colour = "black"),
axis.text.x = element_text(angle= 30, hjust=1, size= 9),
legend.position = "none")
For further examination, I would like to compare the total number of cases caused by P.falcuparum of these countries to total cases, P.viax and other factors.
plot4 <-
plot3 %>% pivot_longer(cols = c("Total_cases","P.falciparum","P.vivax","Others"),
names_to = "Cases",
values_to = "Count",
values_drop_na = TRUE) %>%
select(- (emalaria :eothers))
plot4$ecount <- formatC(plot4$Count, format = "e", digits = 2)
plot4$Cases <- factor(plot4$Cases, levels = c ("P.falciparum", "Total_cases", "P.vivax", "Others"))
b <- ggplot(plot4, aes(x = reorder(Location, - Count), y = Count, text1 = ecount, fill = Cases)) +
geom_bar(stat = "identity", position = "dodge") +
labs( x = "Country", y= "Number of cases", title = "Top ten countries with P.falciparum causes in 2020") +
# scale_y_continuous(limits = c(0, 2.5e+07), labels = label_scientific()) +
theme(plot.title = element_text(hjust = 0.5)) + # Adjust the title to the center
theme(plot.title = element_text(size = (15), face = "bold" ),
axis.title = element_text(size = (13), face = "bold", colour = "black"),
axis.text.x = element_text(angle= 30, hjust=1, size= 9))
ggplotly (b)Result: Among top ten countries with P.falciparum causes in 2020: + 5/10 countries were lack of data.
4/10 countries (Mozambique, Zambia, Côte d’Ivoire and Kenya) had no P.vivax cases or others.
1/10 country (Tanzania) had no P.vivax cases but other causes.
1.3 The P.vivax cases
For my next figure, I am going to create a bar chart that plots top ten countries with highest malaria cases caused by P.vivax on the x axis and the number of malaria cases the y axis in 2020.
# Select the top 10 countries for P.vivax cases:
plot5 <- data %>% arrange(desc(Period),desc(P.vivax)) %>% slice(1:10)
datatable(plot5,
filter = 'top',
caption = htmltools::tags$caption(
style = 'caption-side: top; text-align: center;',
'Table 1.4: ', htmltools::em('The number of top ten malaria cases caused by *P.vivax*')))# Display the e value:
plot5$emalaria <- formatC(plot5$Total_cases, format = "e", digits = 2) # Note: class of "emalaria" is character, not numeric
plot5$efalciparum <- formatC(plot5$P.falciparum, format = "e", digits = 2)
plot5$evivax <- formatC(plot5$P.vivax, format = "e", digits = 2)
plot5$eothers <- formatC(plot5$Others, format = "e", digits = 2)
# Draw a plot
ggplot(plot5, aes(x = reorder(Location, - P.vivax), y = P.vivax, fill = -P.vivax)) +
geom_bar(stat = "identity", color='blue') +
labs( x = "Country", y= "Number of cases", title = "Top ten countries with highest P.vivax cases in 2020") +
geom_text(aes(label = evivax), vjust = - 0.5, size = 3) +
theme(plot.title = element_text(hjust = 0.5)) + # Adjust the title to the center
theme(plot.title = element_text(size = (15), face = "bold" ),
axis.title = element_text(size = (13), face = "bold", colour = "black"),
axis.text.x = element_text(angle= 30, hjust=1, size= 9),
legend.position = "none")
For further examination, I would like to compare the total number of cases caused by P.falcuparum of these countries to total cases, P.viax and other factors.
plot6 <-
plot5 %>% pivot_longer(cols = c("P.falciparum","P.vivax","Others"),
names_to = "Cases",
values_to = "Count",
values_drop_na = TRUE) %>%
select(- (emalaria :eothers))
plot6$ecount <- formatC(plot6$Count, format = "e", digits = 2)
plot6$Cases <- factor(plot6$Cases, levels = c ("P.vivax", "P.falciparum", "Others"))
c <- ggplot(plot6, aes(x = reorder(Location, -Count), y = Count, fill = Cases)) +
geom_bar(stat = "identity", position = "dodge") +
labs( x = "Country", y= "Number of cases", title = "Top ten countries with P.vivax causes in 2020") +
theme(plot.title = element_text(hjust = 0.5)) + # Adjust the title to the center
theme(plot.title = element_text(size = (15), face = "bold" ),
axis.title = element_text(size = (13), face = "bold", colour = "black"),
axis.text.x = element_text(angle= 30, hjust=1, size= 9))
ggplotly (c)Result: Among top ten countries with P.vivax causes in 2020:
In Papua New Guinea, the number of cases caused by P.falciparum and P.vivax were almost equal. The interesting event was that 3.7e+05 cases caused by other factors which was almost double the cases caused by P.falciparum and P.vivax.
Ethiopia, Sudan, Indonesia and India had more P.falciparum cases than P.vivax cases.
In the remaining 5 countries, the P.vivax caused more cases than P.falciparum.
Part 2: The highest cases vs degree of change of top ten countries over 10 years
For my next figure, I am going to create 2 plots to compare the differences of countries between the highest number of cases with the highest degree of changes through each year.
2.1 Highest cases
In this section, I will plot a bar-chart that illustrates the duration (from 2010 to 2020) on the x axis and the number of malaria cases the y axis of top 3 countries having highest cases in each year.
# Select the top 3 countries for Total cases:
plot7 <-
data %>%
arrange(desc(Period),desc(Total_cases)) %>%
group_by(Period) %>%
slice(1:3)
datatable(plot7,
filter = 'top',
caption = htmltools::tags$caption(
style = 'caption-side: top; text-align: center;',
'Table 1.5: ', htmltools::em('The number of top 5 malaria cases from 2010 to 2020')))#Create a column of character vector that change the way the value is present:
plot7$eTotal_cases <- formatC(plot7$Total_cases, format = "e", digits = 2)
#Reorder the factor
plot7$Location <- factor(plot7$Location, levels = c ("Angola", "Burundi", "Ghana", "Liberia", "Tanzania", "Burkina Faso", "Mozambique","Uganda","Nigeria","Congo"))
#Draw a plot
a <- ggplot(plot7, aes(x = Period, y = Total_cases, label = eTotal_cases, fill = Location)) +
geom_bar(stat = "identity", position = "dodge" ,width = 0.7) +
labs( x = "Year", y= "Number of cases", title = "Top 3 highest malaria cases 2010-2020") +
scale_fill_brewer(palette = "Paired") +
theme(plot.title = element_text(hjust = 0.5)) + # Adjust the title to the center
theme(plot.title = element_text(size = (15), face = "bold" ),
axis.title = element_text(size = (13), face = "bold", colour = "black"),
axis.text.x = element_text(angle= 30, hjust=1, size= 9))
ggplotly(a)Result:
Congo is the country with the highest malaria cases from 2010 to 2020.
From 2014 to 2020: Nigeria has the second highest malaria cases, followed by Uganda or Mozambique or Burkina Faso in the 3rd place.
2.2 Top degree of changes
In this section, I will plot a line-graph that plots the duration (from 2010 to 2020) on the x axis and the number of malaria cases the y axis of top 3 countries having highest changes in each year.
First, I compare the degree of changes in the total number of cases:
# Changing the data to wide format for further calculation (must use the single malaria data)
plot8 <-
malaria %>% select(Location, Period, Total_cases ) %>%
pivot_wider(names_from = Period, values_from = Total_cases, names_glue = "Y{Period}")
#Assign value to 0.1 instead of 0 for division:
plot8[plot8==0] <- 0.1
#Calculate the change in the total number of cases (between 2 Year H1 = 2011-2010):
plot8 <-
plot8 %>% mutate(
"2010-2011" = (Y2011 - Y2010) /Y2010,
"2011-2012" = (Y2012 - Y2011) / Y2011,
"2012-2013" = (Y2013 - Y2012) / Y2012,
"2013-2014" = (Y2014 - Y2013) /Y2013,
"2014-2015" = (Y2015 - Y2014) /Y2014,
"2015-2016" = (Y2016 - Y2015) /Y2015,
"2016-2017" = (Y2017 - Y2016) /Y2016,
"2017-2018" = (Y2018 - Y2017) /Y2017,
"2018-2019" = (Y2019 - Y2018) /Y2018,
"2019-2020" = (Y2020 - Y2019) /Y2019,) %>%
pivot_longer(cols = c("2010-2011" :"2019-2020"),
names_to = "Delta",
values_to = "Proportion",
values_drop_na = TRUE)
#Select only 3 columns Location - Delta - Changes for visualization
plot8 <- plot8 %>%
select(Location, Delta,Proportion) %>%
arrange(desc(Delta),desc(Proportion)) %>%
group_by(Delta) %>%
slice(1:3)
# Add a column called Percentage
plot8$Proportion = round(plot8$Proportion,2)
plot8 = plot8 %>% mutate(Percentage= percent(Proportion))
# Show a table
datatable(plot8,
filter = 'top',
caption = htmltools::tags$caption(
style = 'caption-side: top; text-align: center;',
'Table 1.6: ', htmltools::em('Top 3 increased malaria cases 2010-2020')))#Draw a plot
a <- ggplot(plot8, aes(x = Delta, y = Proportion, text1 = Percentage, fill = Location)) +
geom_bar(stat = "identity", position = "dodge" ,width = 0.7) +
labs( x = "Year", y= "Percentage change", title = "Top 3 increased cases between 2010 and 2020") +
scale_y_continuous(labels = scales::percent, limits=c(0,200))+
theme(plot.title = element_text(hjust = 0.5)) + # Adjust the title to the center
theme(plot.title = element_text(size = (14), face = "bold" ),
axis.title = element_text(size = (13), face = "bold", colour = "black"),
axis.text.x = element_text(angle= 30, hjust=1, size= 9))
ggplotly(a)Result:
South Sudan experienced the highest increase during 2016-2017: 19430%. Chat and Djibouti experience the highest increase during 2012-2013: 9687% and 6636%.
The remaining countries experienced under 5000% increase.
Part 3: The linegraph over year
3.1 Cases
Based on the 2 bar charts drawn in Part 2, I choose 6 countries for line graphs:
3 from highest cases: Congo, Nigeria, Uganda.
3 from highest degree of increase: South Sudan, Chad and Djibouti.
plot9 <-
data %>%
filter (Location %in% c("Congo", "Nigeria", "Uganda", "South Sudan", "Chad", "Djibouti"))
# Show a table
datatable(plot9,
filter = 'top',
caption = htmltools::tags$caption(
style = 'caption-side: top; text-align: center;',
'Table 1.7: ', htmltools::em( 'Six countries with malaria cases 2010-2020')))#Draw a plot
ggplot(plot9, aes(x = Period, y = Total_cases, colour = Location)) +
stat_summary(geom = "point") + stat_summary(geom = "line", aes(group = Location)) +
labs( x = "Year", y= "Total cases", title = "Six countries with malaria cases 2010-2020") +
theme(plot.title = element_text(hjust = 0.5)) + # Adjust the title to the center
theme(plot.title = element_text(size = (14), face = "bold" ),
axis.title = element_text(size = (13), face = "bold", colour = "black"),
axis.text.x = element_text(angle= 30, hjust=1, size= 9)) 
3.2 Degree of changes
Similar 6 countries are chosen to plot. Below is the table and the graph based on the proportion of change:
## Extract data similar to plot8 section
plot10 <-
malaria %>%
filter (Location %in% c("Congo", "Nigeria", "Uganda", "South Sudan", "Chad", "Djibouti"))%>% select(Location, Period, Total_cases ) %>%
pivot_wider(names_from = Period, values_from = Total_cases, names_glue = "Y{Period}")
#Assign value to 0.1 instead of 0 for division:
plot10[plot10==0] <- 0.1
#Calculate the change in the total number of cases (between 2 Year H1 = 2011-2010):
plot10 <-
plot10 %>% mutate(
"2010-2011" = (Y2011 - Y2010) /Y2010,
"2011-2012" = (Y2012 - Y2011) / Y2011,
"2012-2013" = (Y2013 - Y2012) / Y2012,
"2013-2014" = (Y2014 - Y2013) /Y2013,
"2014-2015" = (Y2015 - Y2014) /Y2014,
"2015-2016" = (Y2016 - Y2015) /Y2015,
"2016-2017" = (Y2017 - Y2016) /Y2016,
"2017-2018" = (Y2018 - Y2017) /Y2017,
"2018-2019" = (Y2019 - Y2018) /Y2018,
"2019-2020" = (Y2020 - Y2019) /Y2019,) %>%
pivot_longer(cols = c("2010-2011" :"2019-2020"),
names_to = "Delta",
values_to = "Proportion",
values_drop_na = TRUE)
# Select only 3 columns Location - Delta - Changes for visualization
plot10 <- plot10 %>%
select(Location, Delta,Proportion)
# Show a table
datatable(plot10,
filter = 'top',
caption = htmltools::tags$caption(
style = 'caption-side: top; text-align: center;',
'Table 1.8: ', htmltools::em('Degree of malaria changes in Six countries 2010-2020')))#Draw a plot
a <- ggplot(plot10, aes(x = Delta, y = Proportion, colour = Location)) +
stat_summary(geom = "point") + stat_summary(geom = "line", aes(group = Location)) +
labs( x = "Year interval", y= "Proportion change", title = "Degree of changes of 6 countries 2010-2020") +
scale_y_continuous(labels = scales::percent, limits=c(-1,200))+
theme(plot.title = element_text(hjust = 0.5)) + # Adjust the title to the center
theme(plot.title = element_text(size = (14), face = "bold" ),
axis.title = element_text(size = (13), face = "bold", colour = "black"),
axis.text.x = element_text(angle= 30, hjust=1, size= 9))
ggplotly(a) Result:
Among 6 countries, only South Sudan (2016-2017), Djibouti and Chad (2012-2013) experienced a sudden increase to over 5000%.
Besides, between 2011-2012 and 2018-2019, Uganda and South Sudan also experienced an increase over 1000%.
In the remaining period, these 6 countries only suffered a degree of change between -95%-460%
Part 4: The boxplot of different countries over 10 years
4.1 Malaria cases
plot11 <-
data %>% select("Location", "Period","Total_cases") %>%
filter(is.na(Total_cases) == FALSE) %>%
ggplot(aes(y=Total_cases)) +
geom_boxplot(aes(frame = Location)) +
theme(axis.ticks.x = element_blank(),
axis.text.x = element_blank()) +
labs(x ="Location", y = "Total cases", title = "The malaria cases of each country between 2010 and 2020")
# Saving:
## Sys.setenv("plotly_username"="Minh_Tri_92")
## Sys.setenv("plotly_api_key"="B6mso8qyGEpF7kJekgKe")
## api_create(plot11, "The malaria cases of each country between 2010 and 2020")
ggplotly(plot11)4.2 P.falciparum cases
plot11 <-
data %>% select("Location", "Period","P.falciparum") %>%
filter(is.na(P.falciparum) == FALSE) %>%
ggplot(aes(y=P.falciparum)) +
geom_boxplot(aes(frame = Location)) +
theme(axis.ticks.x = element_blank(),
axis.text.x = element_blank()) +
labs(x ="Location", y = "Total cases", title = "The malaria cases caused by P.falciparum between 2010 and 2020")
ggplotly(plot11)4.3 P.vivax cases
plot11 <-
data %>% select("Location", "Period","P.vivax") %>%
filter(is.na(P.vivax) == FALSE) %>%
ggplot(aes(y=P.vivax)) +
geom_boxplot(aes(frame = Location)) +
theme(axis.ticks.x = element_blank(),
axis.text.x = element_blank()) +
labs(x ="Location", y = "Total cases", title = "The malaria cases caused by P.vivax between 2010 and 2020")
ggplotly(plot11)