#Step 1
#install libraries if not installed yet
install.packages("RCurl")
install.packages("RJSONIO")
install.packages("sqldf")
install.packages("jsonlite")
#Call the libraries
library("RCurl")
library("RJSONIO")
library("sqldf")
library("jsonlite")

link<-'https://data.maryland.gov/api/views/pdvh-tf2u/rows.json?accessType=DOWNLOAD'
JSONInput<-fromJSON(txt=link)

JSONData<-JSONInput$data
MyData<-data.frame(JSONData,stringsAsFactors = FALSE)

#Step 2 Cleansing
str(MyData)

MyData<- MyData[,-1:-8]

namesOfColumns<- colnames(MyData)
namesOfColumns <- c("CASE_NUMBER","BARRACK","ACC_DATE","ACC_TIME","ACC_TIME_CODE","DAY_OF_WEEK","ROAD","INTERSECT_ROAD","DIST_FROM_INTERSECT","DIST_DIRECTION","CITY_NAME","COUNTY_CODE","COUNTY_NAME","VEHICLE_COUNT","PROP_DEST","INJURY","COLLISION_WITH_1","COLLISION_WITH_2")
colnames(MyData)<- namesOfColumns
colnames(MyData)

#Step 3
#Get rid of null

step2<- MyData
step22<-na.omit(step2)
any(is.na(step22))
MyData<- step22

# returns string w/o leading or trailing whitespace

MyData$DAY_OF_WEEK <- gsub(" ", "", MyData$DAY_OF_WEEK)
View(MyData)

step3result<- nrow(sqldf("select * from MyData where DAY_OF_WEEK='SUNDAY'"))
step3result

step32result<- sqldf("select count(*) from MyData where INJURY='YES'")
step32result

step33result<- sqldf("select Count(*), DAY_OF_WEEK from MyData Group By DAY_OF_WEEK")
step33result

#Step 4
tapply(MyData$DAY_OF_WEEK, MyData$DAY_OF_WEEK=='SUNDAY', length)
tapply(MyData$INJURY, MyData$INJURY=='YES', length)
tapply(MyData$DAY_OF_WEEK, MyData$DAY_OF_WEEK, length)