R/Overview_DevicePerformanceAnalysis.R

In erenturunc/AnalyticsMiner: What the package does (short line)

devicePerformanceAnalysis <- function(ClientID, ClientSecret, ViewID, StartDate, EndDate, SplitDaywise = T){
  
  checkPackages("RGoogleAnalytics","httpuv", "ggplot2", "scales", "sqldf")
  library(httpuv)
  library(ggplot2)
  library(RGoogleAnalytics)
  library(scales)
  library(sqldf)
  
  token <- Auth(ClientID,ClientSecret)
  ValidateToken(token)
  
  ###Device Performance Analysis - START###
  query.list <- Init(start.date = toString(StartDate),
                     end.date = toString(EndDate),
                     dimensions = "ga:date,ga:deviceCategory",
                     metrics = "ga:sessions,ga:uniquePageviews,ga:goalCompletionsAll,ga:uniquePurchases,ga:transactionRevenue",
                     max.results = 10000,
                     sort = "ga:date",
                     table.id = ViewID)
  
  ga.query <- QueryBuilder(query.list)
  
  ga.data <- GetReportData(ga.query, token, split_daywise = SplitDaywise)
  
  ga.data$date <- as.Date(ga.data$date, format = '%Y%m%d')
  
  p_sessions <- ggplot(ga.data, aes(x=date, y=sessions, group = deviceCategory, color=deviceCategory))+ 
    #geom_errorbar(aes(ymin=sessions-15, ymax=sessions+15), width=.1, 
    #              position=position_dodge(0.05)) +
    geom_line() + 
    labs(title="Sessions by date",x="Dates", y = "Sessions")+
    theme_classic()
  p_sessions <- p_sessions + theme(axis.text.x = element_text(angle = 90, hjust = 1))
  p_sessions <- p_sessions + theme_classic()
  p_sessions <- p_sessions + stat_smooth(method = "lm")
  
  p_uniquePageviews <- ggplot(ga.data, aes(x=date, y=uniquePageviews, group = deviceCategory, color=deviceCategory))+ 
    geom_line() + 
    labs(title="Unique PageViews by date",x="Dates", y = "Unique PageViews")+
    theme_classic()
  p_uniquePageviews <- p_uniquePageviews + theme(axis.text.x = element_text(angle = 90, hjust = 1))
  p_uniquePageviews <- p_uniquePageviews + theme_classic()
  p_uniquePageviews <- p_uniquePageviews + stat_smooth(method = "lm")
  
  p_goalCompletionsAll <- ggplot(ga.data, aes(x=date, y=goalCompletionsAll, group = deviceCategory, color=deviceCategory))+ 
    geom_line() + 
    labs(title="Goal completions by date",x="Dates", y = "Goal Completions")+
    theme_classic()
  p_goalCompletionsAll <- p_goalCompletionsAll + theme(axis.text.x = element_text(angle = 90, hjust = 1))
  p_goalCompletionsAll <- p_goalCompletionsAll + theme_classic()
  p_goalCompletionsAll <- p_goalCompletionsAll + stat_smooth(method = "lm")
  
  p_uniquePurchases <- ggplot(ga.data, aes(x=date, y=uniquePurchases, group = deviceCategory, color=deviceCategory))+ 
    geom_line() + 
    labs(title="Purchases by date",x="Dates", y = "Purchase")+
    theme_classic()
  p_uniquePurchases <- p_uniquePurchases + theme(axis.text.x = element_text(angle = 90, hjust = 1))
  p_uniquePurchases <- p_uniquePurchases + theme_classic()
  p_uniquePurchases <- p_uniquePurchases + stat_smooth(method = "lm")
  
  p_transactionRevenue <- ggplot(ga.data, aes(x=date, y=transactionRevenue, group = deviceCategory, color=deviceCategory))+ 
    geom_line() + 
    labs(title="Revenue by date",x="Dates", y = "Revenue")+
    theme_classic()
  p_transactionRevenue <- p_transactionRevenue + theme(axis.text.x = element_text(angle = 90, hjust = 1))
  p_transactionRevenue <- p_transactionRevenue + theme_classic()
  p_transactionRevenue <- p_transactionRevenue + stat_smooth(method = "lm")
  
  multiplot(p_sessions, p_uniquePageviews, p_goalCompletionsAll, p_uniquePurchases, p_transactionRevenue, cols=2)
  
  data_mobile <- sqldf("select * from [ga.data] where deviceCategory == 'mobile'")
  model_mobile <- lm(formula = sessions ~ date, data = data_mobile)
  predictionDate_mobile = data.frame(date=c('20161101','20161102'))
  predictionDate_mobile$date <- as.Date(predictionDate_mobile$date, format = '%Y%m%d')
  predictedValues_mobile = predict(model_mobile, predictionDate_mobile)
  
  
  ###Device Performance Analysis - END###
  
}