Project2
N Obieyisi and N Nedd
March 9, 2017
- Gini Values Worldwide
- Read in the data, which was downloaded from the reference and uploaded to github
- Select only the Country Name along with the GINI values from 2000 to 2014
- Change to long format, omit N/A values and sort
- Display trend by country using sparkbars embedded in a table
- Summarise by country and find mean for each country
- Map of World Wide mean gini coefficient
- Consumer Complaints Data
- Read in Data
- Find the number of sub-product complaints
- Find the number of sub-product complaints
- Find the product that gets most complaints
- Find the product that gets most complaints with sub products
- TOP 10 Companies with the least timely response and most complaints
- Which state has the most complaints, top 10
- Analyze complaints on a monthly basis per product
- Peform trend analysis compare last years Number of complaints to this years per states with most complaints
- Child Mortality Rates
- Load data with Child mortality rate data
- Get the median values for each country as the value for analysis
- Lets tidy the data and find Infant mortality so that it can be easily analyzed using gather()
- Analyze some countries with high Mortality rates and notice trend in two year periods
- Infant Number of Deaths
- Under 5 mortality analysis
- Neonatal Deaths
Gini Values Worldwide
According to Wikipedia the Gini Coefficient is “a measure of statistical dispersion intended to represent the income or wealth distribution of a nation’s residents, and is the most commonly used measure of inequality.” [https://en.wikipedia.org/wiki/Gini_coefficient]. The data was collected from World Bank’s website at [http://databank.worldbank.org/data/reports.aspx?source=2&series=SI.POV.GINI&country=#]
The Gini Coefficient is given in percent form with 100 being perfect inequality and 0 being perfect equality.
Read in the data, which was downloaded from the reference and uploaded to github
GiniData <- read.csv("https://raw.githubusercontent.com/NNedd/MSDA-Github-repository/master/Data607/Project2/Data_Extract_From_World_Development_Indicators/49d6d15c-fab2-4acf-b212-5d1c8222b8b8_Data.csv", stringsAsFactors = FALSE, na.strings = c("", ".."))
head(GiniData)## ï..Series.Name Series.Code Country.Name Country.Code
## 1 GINI index (World Bank estimate) SI.POV.GINI Afghanistan AFG
## 2 GINI index (World Bank estimate) SI.POV.GINI Albania ALB
## 3 GINI index (World Bank estimate) SI.POV.GINI Algeria DZA
## 4 GINI index (World Bank estimate) SI.POV.GINI American Samoa ASM
## 5 GINI index (World Bank estimate) SI.POV.GINI Andorra ADO
## 6 GINI index (World Bank estimate) SI.POV.GINI Angola AGO
## X1990..YR1990. X2000..YR2000. X2001..YR2001. X2002..YR2002.
## 1 NA NA NA NA
## 2 NA NA NA 31.74
## 3 NA NA NA NA
## 4 NA NA NA NA
## 5 NA NA NA NA
## 6 NA 51.96 NA NA
## X2003..YR2003. X2004..YR2004. X2005..YR2005. X2006..YR2006.
## 1 NA NA NA NA
## 2 NA NA 30.6 NA
## 3 NA NA NA NA
## 4 NA NA NA NA
## 5 NA NA NA NA
## 6 NA NA NA NA
## X2007..YR2007. X2008..YR2008. X2009..YR2009. X2010..YR2010.
## 1 NA NA NA NA
## 2 NA 29.98 NA NA
## 3 NA NA NA NA
## 4 NA NA NA NA
## 5 NA NA NA NA
## 6 NA 42.72 NA NA
## X2011..YR2011. X2012..YR2012. X2013..YR2013. X2014..YR2014.
## 1 NA NA NA NA
## 2 NA 28.96 NA NA
## 3 NA NA NA NA
## 4 NA NA NA NA
## 5 NA NA NA NA
## 6 NA NA NA NA
## X2015..YR2015. X2016..YR2016.
## 1 NA NA
## 2 NA NA
## 3 NA NA
## 4 NA NA
## 5 NA NA
## 6 NA NASelect only the Country Name along with the GINI values from 2000 to 2014
GiniData_part <- select(GiniData, Country.Code, Country.Name, X2000..YR2000.:X2014..YR2014.)
head(GiniData_part)## Country.Code Country.Name X2000..YR2000. X2001..YR2001. X2002..YR2002.
## 1 AFG Afghanistan NA NA NA
## 2 ALB Albania NA NA 31.74
## 3 DZA Algeria NA NA NA
## 4 ASM American Samoa NA NA NA
## 5 ADO Andorra NA NA NA
## 6 AGO Angola 51.96 NA NA
## X2003..YR2003. X2004..YR2004. X2005..YR2005. X2006..YR2006.
## 1 NA NA NA NA
## 2 NA NA 30.6 NA
## 3 NA NA NA NA
## 4 NA NA NA NA
## 5 NA NA NA NA
## 6 NA NA NA NA
## X2007..YR2007. X2008..YR2008. X2009..YR2009. X2010..YR2010.
## 1 NA NA NA NA
## 2 NA 29.98 NA NA
## 3 NA NA NA NA
## 4 NA NA NA NA
## 5 NA NA NA NA
## 6 NA 42.72 NA NA
## X2011..YR2011. X2012..YR2012. X2013..YR2013. X2014..YR2014.
## 1 NA NA NA NA
## 2 NA 28.96 NA NA
## 3 NA NA NA NA
## 4 NA NA NA NA
## 5 NA NA NA NA
## 6 NA NA NA NAChange to long format, omit N/A values and sort
giniLong <- gather(GiniData_part,Year, Gini, X2000..YR2000.:X2014..YR2014.)
#giniLong <- na.omit(giniLong)
giniLong_sort <- arrange(giniLong, Country.Name)
giniLong_sort <- na.omit(giniLong_sort)
giniLong_sort$Year <- as.numeric(str_extract(giniLong_sort$Year, "[[:digit:]]{4}"))
head(giniLong_sort)## Country.Code Country.Name Year Gini
## 18 ALB Albania 2002 31.74
## 21 ALB Albania 2005 30.60
## 24 ALB Albania 2008 29.98
## 28 ALB Albania 2012 28.96
## 76 AGO Angola 2000 51.96
## 84 AGO Angola 2008 42.72Display trend by country using sparkbars embedded in a table1
r <- range(giniLong_sort$Gini)
#Setup sparkbars
bar_string <- "type: 'bar', barColor: 'purple', negBarColor: 'green', highlightColor: 'grey'"
cb_bar = JS(paste0("function (oSettings, json) { $('.spark:not(:has(canvas))').sparkline('html', { ",
bar_string, " }); }"), collapse = "")
#Indicate which columns will use sparkbars(the second in this case)
table_columns <- list(list(targets = c(1), render = JS("function(data, type, full){ return '<span class=spark>' + data + '</span>' }")))
trend_details <-
giniLong_sort %>%
group_by(Country.Name) %>%
summarise(Trend = paste(Gini, collapse = ","))
d1 <- datatable(trend_details, rownames = FALSE, options = list(columnDefs = table_columns, fnDrawCallback = cb_bar))
d1$dependencies <- append(d1$dependencies, htmlwidgets:::getDependency("sparkline"))
d1Azerbaijan shows an interesting trend. There was high gini followed by many years of low gini then a spike again
AZE_data <- filter(giniLong_sort, Country.Code == "AZE")
ggplot(AZE_data, aes(Year, Gini)) + geom_col()
Summarise by country and find mean for each country
giniLong_group <- group_by(giniLong_sort, Country.Code)
giniSummary <- summarise(giniLong_group, mean(Gini))
giniSummary_sort <- arrange(giniSummary, `mean(Gini)`)10 Highest/Lowest inequalities
bottom10 <- head(giniSummary_sort,10)
datatable(bottom10, caption = "Lowest Gini Values")top10 <- tail(giniSummary_sort, 10)
datatable(top10, caption = "Highest Gini Values")Based on the tables above the lowest average gini over 2000 to 2014 is Azerbaijan (relative high equality in the country), and the highest is Botswana which signifies high inequality in the country
Map of World Wide mean gini coefficient
#create a map-shaped window
par(mai=c(0,0,0.2,0),xaxs="i",yaxs="i")
#join to a coarse resolution map
spdf <- joinCountryData2Map(giniSummary_sort, joinCode="ISO3", nameJoinColumn="Country.Code")## 144 codes from your data successfully matched countries in the map
## 5 codes from your data failed to match with a country code in the map
## 99 codes from the map weren't represented in your datamapCountryData(spdf, nameColumnToPlot="mean(Gini)")
Consumer Complaints Data
In this result set we will be analyzing the Consumer complaints about financial services
“The Consumer Financial Protection Bureau (CFPB) is the first federal agency solely focused on consumer financial protection and consumer complaints are an integral part of that work. The CFPB helps connect consumers with financial companies to make their voices heard This Monthly Complaint Report provides a high-level snapshot of trends in consumer complaints. The Monthly Complaint Report uses a three-month rolling average, comparing the current average to the same period in the prior year where appropriate, to account for monthly and seasonal fluctuations. In some cases, we use month-to-month comparisons to highlight more immediate trends. For companylevel complaint data, we use a three-month rolling average of complaints sent to companies for response. This company-level complaint data lags other complaint data in this report by two months to reflect the 60 days companies have to respond to complaints, confirming a commercial relationship with the consumer” -https://s3.amazonaws.com/files.consumerfinance.gov/f/documents/201702_cfpb_Monthly-Complaint-Report.pdf
Read in Data
file_path <- c("https://raw.githubusercontent.com/nobieyi00/CUNY_MSDA_R/master/consumer_complaints.csv")
cc_df <- read.csv(file_path, header = TRUE, stringsAsFactors = FALSE,
na.strings = c("N/A",''))Find the number of sub-product complaints
cc_df %>%
group_by(Sub.product) %>%
summarise(count=n())## # A tibble: 38 × 2
## Sub.product count
## <chr> <int>
## 1 (CD) Certificate of deposit 12
## 2 Auto 11
## 3 Cashing a check without an account 2
## 4 Checking account 231
## 5 Conventional adjustable mortgage (ARM) 100
## 6 Conventional fixed mortgage 308
## 7 Credit card 133
## 8 Debt settlement 2
## 9 Domestic (US) money transfer 13
## 10 Federal student loan 8
## # ... with 28 more rowsFind the number of sub-product complaints
cc_df %>%
group_by(Sub.product) %>%
summarise(count=n()) %>%
filter(Sub.product != '' )%>%
arrange(desc(count))## # A tibble: 37 × 2
## Sub.product count
## <chr> <int>
## 1 Other mortgage 315
## 2 Conventional fixed mortgage 308
## 3 Checking account 231
## 4 Other (i.e. phone, health club, etc.) 207
## 5 Credit card 133
## 6 I do not know 126
## 7 FHA mortgage 114
## 8 Other bank product/service 109
## 9 Conventional adjustable mortgage (ARM) 100
## 10 Medical 87
## # ... with 27 more rowsFind the product that gets most complaints
cc_df %>%
group_by(Product) %>%
summarise(No_Products=n()) %>%
arrange(No_Products)## # A tibble: 11 × 2
## Product No_Products
## <chr> <int>
## 1 Other financial service 4
## 2 Prepaid card 18
## 3 Money transfers 27
## 4 Payday loan 28
## 5 Student loan 44
## 6 Consumer Loan 159
## 7 Credit card 339
## 8 Bank account or service 383
## 9 Debt collection 625
## 10 Credit reporting 638
## 11 Mortgage 921Find the product that gets most complaints with sub products
cc_df %>%
filter(Sub.product != '' )%>%
group_by(Product, Sub.product) %>%
summarise(No_Products=n()) %>%
arrange(No_Products)## Source: local data frame [39 x 3]
## Groups: Product [9]
##
## Product Sub.product No_Products
## <chr> <chr> <int>
## 1 Consumer Loan Title loan 1
## 2 Prepaid card Gift or merchant card 1
## 3 Prepaid card Payroll card 1
## 4 Student loan Federal student loan servicing 1
## 5 Bank account or service Cashing a check without an account 2
## 6 Other financial service Debt settlement 2
## 7 Other financial service Refund anticipation check 2
## 8 Prepaid card ID prepaid card 2
## 9 Prepaid card Mobile wallet 2
## 10 Prepaid card Government benefit payment card 3
## # ... with 29 more rowsTOP 10 Companies with the least timely response and most complaints
company_df <- cc_df %>%
filter(Timely.response.=='No')%>%
group_by(Company)%>%
summarise(No_complaints = n())%>%
arrange(desc(No_complaints))
company_ten <-head(company_df , 10)
company_ten## # A tibble: 10 × 2
## Company No_complaints
## <chr> <int>
## 1 Wells Fargo & Company 11
## 2 High Point Asset Inc 3
## 3 Accelerated Receivables Management, Inc. 2
## 4 B.C. Services, Inc. 2
## 5 Bank of the West 2
## 6 FirstBank of Puerto Rico 2
## 7 Mobiloans, LLC 2
## 8 Planet Home Lending, LLC 2
## 9 Real Time Resolutions 2
## 10 Alorica Inc. 1# Generate data
ggplot(company_ten, aes(x=Company,y=No_complaints,fill=Company)) + geom_bar(stat="identity", show.legend = FALSE) + coord_flip()
We clearly see it is Wells fargo is at fault
Which state has the most complaints, top 10
top10_states<- cc_df %>%
group_by(State)%>%
summarise(No_complaints = n())%>%
arrange(desc(No_complaints))%>%
select(State)%>%
head(10)
top10_states## # A tibble: 10 × 1
## State
## <chr>
## 1 CA
## 2 FL
## 3 TX
## 4 NY
## 5 GA
## 6 NJ
## 7 PA
## 8 VA
## 9 MD
## 10 OHAnalyze complaints on a monthly basis per product
Product_Month_df <- cc_df %>%
separate(Date.received, c("Month","Day","Year"), sep = "/") %>%
mutate(Month_num= as.numeric(Month)) %>%
group_by(Product,Month_num)%>%
summarise(No_complaints = n())%>%
arrange(Product , Month_num)
pm_b <-Product_Month_df%>%
filter(Product=="Bank account or service")
ggplot(Product_Month_df, aes(Month_num, No_complaints,colour=Product)) +
geom_line() +
geom_point() +ggtitle(label="Trend of complaints per month") 
We notice that Debt collection is usually high in beginning of year
plot(pm_b$Month_num, pm_b$No_complaints)
lines(pm_b$Month_num[order(pm_b$Month_num)], pm_b$No_complaints)
title(main="Bank Accout or Service trend per month", xlab="Month", ylab="Number of complaints")
Peform trend analysis compare last years Number of complaints to this years per states with most complaints
S_y_complain <-cc_df %>%
separate(Date.received, c("Month","Day","Year"), sep = "/") %>%
mutate(Year_num= as.numeric(Year)) %>%
group_by(State,Year_num)%>%
summarise(No_complaints = n())%>%
filter(State %in% top10_states$State )%>%
arrange(State,Year_num)
S_y_complain ## Source: local data frame [27 x 3]
## Groups: State [10]
##
## State Year_num No_complaints
## <chr> <dbl> <int>
## 1 CA 2013 58
## 2 CA 2016 375
## 3 CA 2017 4
## 4 FL 2013 37
## 5 FL 2016 251
## 6 FL 2017 1
## 7 GA 2013 18
## 8 GA 2016 156
## 9 GA 2017 1
## 10 MD 2013 3
## # ... with 17 more rowsggplot(S_y_complain, aes(x=State,y=No_complaints,fill=Year_num)) + geom_bar(stat="identity", show.legend = TRUE) 
Child Mortality Rates
Child mortality rate around the world is a bench mark used to know how healthy are the childeren around the world. World leaders need to pay attention to their countries child mortality to avoid catastrophic consequences in future generation. In this data set we would study the child mortality rates This data was retrieved from http://www.childmortality.org/files_v20/download/RatesDeaths_AllIndicators.xlsx
Load data with Child mortality rate data
file_path <- c("https://raw.githubusercontent.com/nobieyi00/CUNY_MSDA_R/master/childmortality.csv")
CMR_df <- read.csv(file_path, header = TRUE, stringsAsFactors = FALSE,
na.strings = c("N/A",''))Get the median values for each country as the value for analysis
CMR_1<-CMR_df %>%
filter(Uncertainty.bounds. == 'Median')Lets tidy the data and find Infant mortality so that it can be easily analyzed using gather()
IM <-CMR_1 %>%
gather("type_year","Value",4:ncol(CMR_1)) %>%
#filter(type_year %in% c('Infant.Deaths.','IMR'))%>%
mutate(year= as.numeric(str_sub(type_year,-4,-1)), type = str_sub(type_year,1,str_length(type_year)-5) )%>%
filter(type %in% c('Infant.Deaths','IMR'))%>%
filter(is.na(Value)==FALSE)%>%
select (CountryName, year,type,Value)Analyze some countries with high Mortality rates and notice trend in two year periods
Cnt_10<-IM%>%
filter(type=="IMR" )%>%
mutate(year_range=ifelse(year %in% c(1950:1970), '1950_1970',
ifelse(year %in% c(1971:1990),'1971_1990',
ifelse(year %in% c(1991:2010),'1991_2010',
ifelse(year %in% c(2011:2030),'2011_2030',NA)))))%>%
group_by(CountryName,year_range) %>%
summarise(avg = sum(Value)/n())%>%
arrange(desc(avg))%>%
select(CountryName)%>%
head(10)
tre<- IM%>%
filter(type=="IMR" )%>%
filter(CountryName %in% Cnt_10$CountryName )%>%
mutate(year_range=ifelse(year %in% c(1950:1970), '1950_1970',
ifelse(year %in% c(1971:1990),'1971_1990',
ifelse(year %in% c(1991:2010),'1991_2010',
ifelse(year %in% c(2011:2030),'2011_2030',NA)))))%>%
group_by(CountryName,year_range) %>%
summarise(avg = sum(Value)/n())%>%
arrange(CountryName)
ggplot(tre, aes(x=CountryName,y=avg,fill=year_range)) + geom_bar(stat="identity", show.legend = TRUE) 
Cnt_10<-IM%>%
filter(type=="IMR" )%>%
mutate(year_range=ifelse(year %in% c(1950:1970), '1950_1970',
ifelse(year %in% c(1971:1990),'1971_1990',
ifelse(year %in% c(1991:2010),'1991_2010',
ifelse(year %in% c(2011:2030),'2011_2030',NA)))))%>%
group_by(CountryName,year_range) %>%
summarise(avg = sum(Value)/n())%>%
arrange(desc(avg))%>%
select(CountryName)%>%
head(3)
tre<- IM%>%
filter(type=="IMR" )%>%
filter(CountryName %in% Cnt_10$CountryName )%>%
mutate(year_range=ifelse(year %in% c(1950:1970), '1950_1970',
ifelse(year %in% c(1971:1990),'1971_1990',
ifelse(year %in% c(1991:2010),'1991_2010',
ifelse(year %in% c(2011:2030),'2011_2030',NA)))))%>%
group_by(CountryName,year_range) %>%
summarise(avg = sum(Value)/n())%>%
arrange(CountryName)
ggplot(tre, aes(year_range, avg,colour=CountryName, group=CountryName
)) + geom_point() +ggtitle(label="Trend of infant mortality rates over the years") +
geom_line()
Infant Number of Deaths
Lets look the country that had the most reduced rate of child deaths over time. This country we can say made the most improvement
But first lets look at any country and plot it out over time
france_ND <-IM%>%
filter(type=="Infant.Deaths" & CountryName== "France" )
ggplot(france_ND, aes(year, Value,colour=CountryName, group=CountryName
)) + geom_point() +ggtitle(label="Trend of Infant number of deaths over the years") +
geom_line()
#find coorelation
corr_france <- cor(france_ND$Value,france_ND$year)
sd(france_ND$year)## [1] 19.19635sd(france_ND$Value)## [1] 9336.837Regression_slop <-corr_france*(sd(france_ND$Value)/sd(france_ND$year))Lets look at the trend of Zambia
Zambia_ND <-IM%>%
filter(type=="Infant.Deaths" & CountryName== "Zambia" )
ggplot(Zambia_ND, aes(year, Value,colour=CountryName, group=CountryName
)) + geom_point() +ggtitle(label="Trend of Infant number of deaths over the years") +
geom_line()
#find coorelation
corr_Zambia <- cor(Zambia_ND$Value,Zambia_ND$year)
sd(Zambia_ND$year)## [1] 18.3303sd(Zambia_ND$Value)## [1] 9402.08Regression_slop <-corr_Zambia*(sd(Zambia_ND$Value)/sd(Zambia_ND$year))
Regression_slop## [1] 379.2305Lets look at Nigeria
Nigeria_ND <-IM%>%
filter(type=="Infant.Deaths" & CountryName== "Nigeria" )
ggplot(Nigeria_ND, aes(year, Value,colour=CountryName, group=CountryName
)) + geom_point() +ggtitle(label="Trend of Infant number of deaths over the years") +
geom_line()
##What is regression slope for all the countries ###Lets get the top 10 countries
IM%>%
filter(type=="Infant.Deaths" )%>%
group_by(CountryName)%>%
summarize(Cor_co = cor(Value,year), sd_year=sd(year), sd_value=sd(Value) ,Regression_slope=cor(Value,year)*(sd(Value)/sd(year)))%>%
arrange(Regression_slope)%>%
head(10)## Warning in cor(c(16874, 17437, 17478, 16961, 16345, 15696, 15158, 14891, :
## the standard deviation is zero
## Warning in cor(c(16874, 17437, 17478, 16961, 16345, 15696, 15158, 14891, :
## the standard deviation is zero## # A tibble: 10 × 5
## CountryName Cor_co sd_year sd_value Regression_slope
## <chr> <dbl> <dbl> <dbl> <dbl>
## 1 China -0.8917827 13.71131 640440.74 -41654.224
## 2 India -0.9458409 18.33030 674350.51 -34796.388
## 3 Indonesia -0.9812851 18.90767 166591.65 -8645.904
## 4 Brazil -0.9879252 17.75293 130914.41 -7285.198
## 5 Bangladesh -0.8016545 18.33030 116985.35 -5116.219
## 6 Egypt -0.9770570 17.75293 79095.60 -4353.134
## 7 Turkey -0.9913345 18.33030 70690.77 -3823.079
## 8 Iran -0.9697807 13.13393 47994.76 -3543.830
## 9 Mexico -0.9777475 15.44345 53748.73 -3402.912
## 10 Thailand -0.9758883 19.19635 43217.27 -2197.043Lets look at China
China_ND <-IM%>%
filter(type=="Infant.Deaths" & CountryName== "China" )
ggplot(China_ND, aes(year, Value,colour=CountryName, group=CountryName
)) + geom_point() +ggtitle(label="Trend of Infant number of deaths over the years") +
geom_line()
###Lets get the bottom 10 countries
IM%>%
filter(type=="Infant.Deaths" )%>%
group_by(CountryName)%>%
summarize(Cor_co = cor(Value,year), sd_year=sd(year), sd_value=sd(Value) ,Regression_slope=cor(Value,year)*(sd(Value)/sd(year)))%>%
arrange(desc(Regression_slope))%>%
head(10)## Warning in cor(c(16874, 17437, 17478, 16961, 16345, 15696, 15158, 14891, :
## the standard deviation is zero
## Warning in cor(c(16874, 17437, 17478, 16961, 16345, 15696, 15158, 14891, :
## the standard deviation is zero## # A tibble: 10 × 5
## CountryName Cor_co sd_year sd_value Regression_slope
## <chr> <dbl> <dbl> <dbl> <dbl>
## 1 Nigeria 0.8135546 15.15476 59767.937 3208.5290
## 2 Congo DR 0.9606360 13.71131 40002.844 2802.6623
## 3 Angola 0.9765648 10.53565 17618.692 1633.1017
## 4 Uganda 0.7131499 18.04162 20326.875 803.4816
## 5 Chad 0.9927706 12.84523 9811.845 758.3289
## 6 Sudan 0.9040948 18.33030 13998.715 690.4504
## 7 Niger 0.8290873 14.28869 9388.198 544.7410
## 8 Tanzania 0.5495608 18.90767 17787.294 516.9965
## 9 Cote d Ivoire 0.8912637 17.17556 9685.933 502.6164
## 10 Somalia 0.9004450 9.66954 4583.354 426.8102This shows Nigeria infant deaths increased over time and is the worst in the world. We can also say that the top 10 infant child deaths are in African countries. However, a plot above shows it is is reducing. I am from Nigeria and I am surprised that our infant mortality rose the highest.
Under 5 mortality analysis
Under5_data <- select(CMR_df, ISO.Code:U5MR.2015,Under.five.Deaths.1950:Under.five.Deaths.2015)
datatable(head(Under5_data))Under 5 mortality rate analysis
Under5_rate <- Under5_data %>% select(ISO.Code:U5MR.2015) %>%
filter(Uncertainty.bounds. == "Median")Transform data into long form
under5_rateLong <- gather(Under5_rate,Year, MortalityRate, U5MR.1950:U5MR.2015)
under5_rateLong$Year <- as.numeric(str_extract(under5_rateLong$Year, "[[:digit:]]{4}"))
head(under5_rateLong)## ISO.Code CountryName Uncertainty.bounds. Year MortalityRate
## 1 AFG Afghanistan Median 1950 NA
## 2 AGO Angola Median 1950 NA
## 3 ALB Albania Median 1950 NA
## 4 AND Andorra Median 1950 NA
## 5 ARE United Arab Emirates Median 1950 NA
## 6 ARG Argentina Median 1950 NAEliminate N/A values and redundant Uncertainty Bounds Column
Under5_rateSome <- na.omit(under5_rateLong)
Under5_rateSome <- select(Under5_rateSome, ISO.Code, CountryName, Year, MortalityRate)Display Map of Mortality rates
#create a map-shaped window
par(mai=c(0,0,0.2,0),xaxs="i",yaxs="i")
#join to a coarse resolution map
spdf <- joinCountryData2Map(Under5_rateSome, joinCode="ISO3", nameJoinColumn="ISO.Code")## 10244 codes from your data successfully matched countries in the map
## 0 codes from your data failed to match with a country code in the map
## 48 codes from the map weren't represented in your datamapCountryData(spdf, nameColumnToPlot="MortalityRate",missingCountryCol = "grey",colourPalette = c("green", "yellow", "red"))## Warning in rwmGetColours(colourPalette, numColours): 3 colours specified
## and 7 required, using interpolation to calculate colours
From the map it can be seen that countries in South America, Africa and Asia tend to have higher mortality rates than the rest of the world
One exeption in Africa is Namibia
namibia_data <- filter(Under5_rateSome, CountryName == "Namibia")
boxplot(namibia_data$MortalityRate)
Isolate Numbers
Under5_count <- Under5_data %>% select(ISO.Code:Uncertainty.bounds.,Under.five.Deaths.1950:Under.five.Deaths.2015) %>%
filter(Uncertainty.bounds. == "Median")Transform data into long form
under5_countLong <- gather(Under5_count,Year, MortalityRate, Under.five.Deaths.1950:Under.five.Deaths.2015)
under5_countLong$Year <- as.numeric(str_extract(under5_countLong$Year, "[[:digit:]]{4}"))
head(under5_countLong)## ISO.Code CountryName Uncertainty.bounds. Year MortalityRate
## 1 AFG Afghanistan Median 1950 NA
## 2 AGO Angola Median 1950 NA
## 3 ALB Albania Median 1950 NA
## 4 AND Andorra Median 1950 NA
## 5 ARE United Arab Emirates Median 1950 NA
## 6 ARG Argentina Median 1950 NAEliminate N/A values and redundant Uncertainty Bounds Column
Under5_countLong <- na.omit(under5_countLong)
Under5_countLong <- select(Under5_countLong, ISO.Code, CountryName, Year, MortalityRate)
head(Under5_countLong)## ISO.Code CountryName Year MortalityRate
## 9 AUS Australia 1950 6169
## 14 BEN Benin 1950 27932
## 15 BFA Burkina Faso 1950 72277
## 30 CAN Canada 1950 17848
## 31 CHE Switzerland 1950 3365
## 49 DNK Denmark 1950 2917r <- range(Under5_countLong$MortalityRate)
#Setup sparklines
box_string <- "type: 'box', lineColor: 'black', whiskerColor: 'black', outlierFillColor: 'black', outlierLineColor: 'black', medianColor: 'black', boxFillColor: 'orange', boxLineColor: 'black'"
cb_box = JS(paste0("function (oSettings, json) { $('.spark:not(:has(canvas))').sparkline('html', { ",
box_string, ", chartRangeMin: ", r[1], ", chartRangeMax: ", r[2], " }); }"),
collapse = "")
#Indicate which columns will use sparklines (the second in this case)
table_columns2 <- list(list(targets = c(1), render = JS("function(data, type, full){ return '<span class=spark>' + data + '</span>' }")))
trend_details2 <-
Under5_countLong %>%
group_by(CountryName) %>%
summarise(Trend = paste(MortalityRate, collapse = ","))
d2 <- datatable(trend_details2, rownames = FALSE, options = list(columnDefs = table_columns2, fnDrawCallback = cb_box))
d2$dependencies <- append(d2$dependencies, htmlwidgets:::getDependency("sparkline"))
d2China and India show great variability in their mortality numbers, almost all other countries show very little variablity in the numbers
china_data <- filter(Under5_countLong, CountryName == "China")
f <- ggplot(china_data, aes(Year,MortalityRate))
f+geom_col()
in the 1900’s China hand a large number of deaths of children under 5. This amount has decreased over the years.
Neonatal Deaths
Lets tidy the data and find Neonatal mortality so that it can be easily analyzed using gather()
NM <-CMR_1 %>%
gather("type_year","Value",4:ncol(CMR_1)) %>%
#filter(type_year %in% c('Infant.Deaths.','IMR'))%>%
mutate(year= as.numeric(str_sub(type_year,-4,-1)), type = str_sub(type_year,1,str_length(type_year)-5) )%>%
filter(type %in% c('Neonatal.Deaths','NMR'))%>%
filter(is.na(Value)==FALSE)%>%
select (ISO.Code,CountryName, year,type,Value)Analyse the trend of those with highest rates
TopNM<-NM%>%
filter(type=="NMR" )%>%
mutate(year_range=ifelse(year %in% c(1950:1970), '1950_1970',
ifelse(year %in% c(1971:1990),'1971_1990',
ifelse(year %in% c(1991:2010),'1991_2010',
ifelse(year %in% c(2011:2030),'2011_2030',NA)))))%>%
group_by(CountryName,year_range) %>%
summarise(avg = sum(Value)/n())%>%
arrange(desc(avg))%>%
select(CountryName)%>%
head(6)
tre<- NM%>%
filter(type=="NMR" )%>%
filter(CountryName %in% TopNM$CountryName )%>%
mutate(year_range=ifelse(year %in% c(1950:1970), '1950_1970',
ifelse(year %in% c(1971:1990),'1971_1990',
ifelse(year %in% c(1991:2010),'1991_2010',
ifelse(year %in% c(2011:2030),'2011_2030',NA)))))%>%
group_by(CountryName,year_range) %>%
summarise(avg = sum(Value)/n())%>%
arrange(CountryName)
ggplot(tre, aes(year_range, avg,colour=CountryName, group=CountryName
)) + geom_point() +ggtitle(label="Trend of neonatal mortality rates over the years") +
geom_line()
###World View
NM_map <- NM%>%
filter(type=="NMR" )
#create a map-shaped window
par(mai=c(0,0,0.2,0),xaxs="i",yaxs="i")
#join to a coarse resolution map
spdf <- joinCountryData2Map(NM_map, joinCode="ISO3", nameJoinColumn="ISO.Code")## 8494 codes from your data successfully matched countries in the map
## 0 codes from your data failed to match with a country code in the map
## 48 codes from the map weren't represented in your datamapCountryData(spdf, nameColumnToPlot="Value",missingCountryCol = "grey",colourPalette = "diverging", mapTitle = "World wide neonatal mortality rates")
This code was adapted from the following two sources: https://leonawicz.github.io/HtmlWidgetExamples/ex_dt_sparkline.html and https://github.com/htmlwidgets/sparkline/issues/3 both accessed on March 12, 2017.↩