|IS 608 KNOWLEDGE AND VISUAL ANALYTICS - PROJECT PROPOSAL | Data Analytics

Aim And Methodology:

• Aim:

The goal of this project is to:

• Show the merit and demerit of visual analytics on data analysis, and how to improve it.
• Provide answer to the question, “Does demographics like Age, Sex, Location have impact on types of death, economic status?”

• Make the visualized image(s) tells the story in an unambiguous way, understandable, even to a layman!

• Methodology:

The project would lay more emphasis on the explanatory techniques. It will be used in making data presentation to the viewers in a more succinct way. I therefore plan to use the R programing language to explore and analysis the dataset.

The dataset to be used is the World Health Nutrition and Population Statistics from year 2010 to 2016 . This can be obtained from http://databank.worldbank.org/data/reports.aspx?source=health-nutrition-and-population-statistics#advancedDownloadOptions .

Load this libraries and dataset and lets get to work!

Glimpse

• Although the dataset contain lots of missing values, we will deal with it!

suppressMessages(library(knitr))
suppressMessages(library(dplyr))
suppressMessages(library(ggplot2))
suppressMessages(library(plotly))
suppressMessages(library(sqldf))
suppressPackageStartupMessages(library(googleVis))

df <- read.csv("WorldHealth.csv", header = TRUE, sep = ",", stringsAsFactors = FALSE)
kable(head(df[200:206, ]))

• World longitudes and latitudes

lat_long <- read.csv("Countries_long_lat2.csv", header = TRUE, sep = ",")
colnames(lat_long) <- c("Country", "Country_Code", "Latitude", "Longtitude")
kable(head(lat_long))

Cleaning and renaming of dataset and column respectively.

options(warn = -1)
df2 <- merge(df, lat_long, by.x = "Country_Code", by.y = "Country_Code", all = FALSE)

df2[, 4:19] <- sapply(df2[, 4:19], as.numeric) # Covet the columns to Numeric

kable(head(df2))

Merging column lonitude and Latitude together for a better coordinate to be in maps (googlevis)

df2$Lat_Long = paste(df2$Latitude, df2$Longtitude, sep=":")

kable(head(df2))

we are now to goint make use of sql to subset(query) columns so as to diffentiate between year 2000 and 2010 where the number children orphaned by HIV/AIDS more than 50000.

Twentyfirst_Cen <- sqldf("SELECT Country_Name, Year_Code, Lat_Long, Population_total, (Children_orphaned_by_HIV_AIDS/Population_total)*100 as 'Percentage_Orphaned_byHIV' FROM df2 where Percentage_Orphaned_byHIV >= 2 ORDER BY Percentage_Orphaned_byHIV DESC LIMIT 50" )

• The world map showing the countries where children are orphaned by HIV?AIDS (2000, 2010)

Twenty First century coutries that are still battling with HIV/AIDS related deaths.

Malaria_Inc_Map <- gvisGeoChart(Twentyfirst_Cen, locationvar ="Lat_Long",  hovervar ="Country_Name",sizevar = "Percentage_Orphaned_byHIV", colorvar = "Population_total",
                   options=list(displayMode="Markers", 
                   colorAxis="{colors:['purple', 'red', 'orange', 'grey', 'pink']}",
                   backgroundColor="lightblue"), chartid="Lost_Their_Parents_To_HIV_AIDS")
plot(Malaria_Inc_Map )

• From the map above, we can see that majority of countries where more that 50000 children lost their parents to HIV/AIDS are in southern part of Africa.

A better way to see for youself!

ggplotly(ggplot(Twentyfirst_Cen, aes(x=Percentage_Orphaned_byHIV, y=reorder(Country_Name, +Percentage_Orphaned_byHIV), fill=Year_Code)) +
geom_point(colour="purple", size=2, alpha=.8) +
scale_fill_brewer(palette="Blues", breaks=rev(levels(Twentyfirst_Cen$Year_Code))) + labs(title="Chart of Children Orphaned By HIV/AIDS BY Countries (%)"))

• The chart below depicts the 20-21st Century Countries Where Most Children Are Losing Their Parent To HIV/AIDS

Combo <- gvisComboChart(Twentyfirst_Cen, xvar="Country_Name",
                                     yvar="Percentage_Orphaned_byHIV",
                                   options=list(seriesType="bars",
                                                bar="{groupWidth:'100%'}",
                                                title="Interactive Chart Of Countries And Related HIV/AIDS Orphaned ",
                                                series='{0: {type:"line"}}'),chartid = "ER")
                     
plot(Combo)

The chart shows that majority of the HIV/AIDS related death were rampant in the late Nineteen century than it were in the 20-21th century.

Tools And Packages

Tools To Be Used:

• GoogleVis
• Plotly
• GGplot2
• Sqldf

Packages To Be Used:

• Plotly
• Knitr
• Dplyr
• Plyr
• MASS
• Reshape2
• Ggplot2
• Graphics
• Ggthemes
• GoogleVis

etc