sugar: Sugar for your social data mining.

#' A overlapOutput Function
#'
#' This function return csv files with Smart Overlap
#' @param love Do you love cats? Defaults to TRUE.
#' @keywords printLog
#' @export
#' @examples
#' a <- "1234"
#' cat_function(a)
#'
overlapOutput <- function(toAnalyze,mainPageList,analyzedFrom,analyzedTo, backgroundFrom, backgroundTo, comments="F") {
  postsDetail_sloucene <- data.frame()
  for(mainPage in mainPageList) {
    
    Target <- as.character(subset(dfPages, username==mainPage)$idPage)
    
    
    toAnalyzeID  <- as.vector(as.character(subset(dfPages, username!=mainPage)$idPage))
    
    postsDetail_all <- data.frame()
    overlap <- data.frame()
    overlapPrct <- data.frame()
    
    for(Reference in toAnalyzeID) {
      
      #### nacteni vsech postu
      tbl <- data.frame()
      
      
      for(i in toAnalyze) {
        filenames <- paste0(i, "_posts.csv", sep="")
        tblTmp <- read.csv(filenames)
        tblTmp$X <- NULL
        tbl <- rbind(tbl, tblTmp)
      }
      
      tbl$id <- as.character(tbl$id)
      tbl$date <-  format(as.POSIXlt(tbl$created_time))
      
      tblMainRange <- subset(tbl, date >= analyzedFrom & date <= analyzedTo)
      tblBackgroundRange  <- subset(tbl, date >= backgroundFrom & date <= backgroundTo)
      
      
      tblMainLikes <- data.frame()
      for(inputPage in toAnalyze) {
        cat("\n")
        cat(inputPage)
        filename <- paste0(inputPage, "_likes.csv", sep="")
        tblLikesTmp <- read.csv(filename)
        tblLikesTmp$X <- NULL
        tblLikesTmp$from_id <- as.character(tblLikesTmp$from_id)
        idPage <- as.character(strsplit(as.character(tblLikesTmp[1,]$post_id), "_")[[1]][1])
        
        postForLikes <- subset(tblMainRange, idPage == idPage)
        needPost <- as.vector(postForLikes$id)
        
        tblLikesTmp <- subset(tblLikesTmp, post_id %in% needPost)
        
        if(comments=="T") {
          
          filename <- paste0(inputPage, "_comments.csv", sep="")
          tblCommentsTmp <- read.csv(filename)
          tblCommentsTmp$from_id <- as.character(tblCommentsTmp$from_id )
          tblCommentsTmp <- data.frame(tblCommentsTmp$from_name, tblCommentsTmp$from_id, tblCommentsTmp$post_id)
          colnames(tblCommentsTmp) <- c("from_name", "from_id", "post_id")
          tblCommentsTmp <- subset(tblCommentsTmp, post_id %in% needPost)
          
          tblLikesTmp <- rbind(tblLikesTmp,tblCommentsTmp)
          
          
          cat(length(unique(tblCommentsTmp$from_id)))
          cat("\n")
        }
        
        tblLikesTmp$idPage <- idPage
        tblMainLikes <- rbind(tblMainLikes,tblLikesTmp)
      }
      
      
      tblBackgroundLikes <- data.frame()
      for(inputPage in toAnalyze) {
        cat("\n")
        cat(inputPage)
        filename <- paste0(inputPage, "_likes.csv", sep="")
        tblLikesTmp <- read.csv(filename)
        tblLikesTmp$X <- NULL
        tblLikesTmp$from_id <- as.character(tblLikesTmp$from_id)
        idPage <- as.character(strsplit(as.character(tblLikesTmp[1,]$post_id), "_")[[1]][1])
        
        postForLikes <- subset(tblBackgroundRange, idPage == idPage)
        needPost <- as.vector(postForLikes$id)
        
        tblLikesTmp <- subset(tblLikesTmp, post_id %in% needPost)
        
        if(comments=="T") {
          
          filename <- paste0(inputPage, "_comments.csv", sep="")
          tblCommentsTmp <- read.csv(filename)
          tblCommentsTmp$from_id <- as.character(tblCommentsTmp$from_id )
          tblCommentsTmp <- data.frame(tblCommentsTmp$from_name, tblCommentsTmp$from_id, tblCommentsTmp$post_id)
          colnames(tblCommentsTmp) <- c("from_name", "from_id", "post_id")
          tblCommentsTmp <- subset(tblCommentsTmp, post_id %in% needPost)
          
          tblLikesTmp <- rbind(tblLikesTmp,tblCommentsTmp)
          
          
          cat(length(unique(tblCommentsTmp$from_id)))
          cat("\n")
        }
        
        tblLikesTmp$idPage <- idPage
        tblBackgroundLikes <- rbind(tblBackgroundLikes,tblLikesTmp)
      }
      
      
      ###
      ### Intersects
      ###
      
      
      TargetName <- subset(dfPages, idPage == Target)
      TargetName <- as.character(TargetName$name)
      
      ReferenceName <- subset(dfPages, idPage == Reference)
      ReferenceName <- as.character(ReferenceName$name)
      
      TargetMain <- subset(tblMainLikes, idPage == Target)
      TargetMain <- unique(TargetMain$from_id)
      
      ReferenceMain <- subset(tblMainLikes, idPage == Reference)
      ReferenceMain <- unique(ReferenceMain$from_id)
      
      TargetBackground <- subset(tblBackgroundLikes, idPage == Target)
      TargetBackground <- unique(TargetBackground$from_id)
      
      ReferenceBackground <- subset(tblBackgroundLikes, idPage == Reference)
      ReferenceBackground <- unique(ReferenceBackground$from_id)
      
      unionMain <- unique(TargetMain,ReferenceMain)
      unionBackground <- unique(TargetBackground, ReferenceBackground)
      
      
      JustMyMain <- unique(setdiff(TargetMain,ReferenceMain))
      JustCompetMain <- unique(setdiff(ReferenceMain,TargetMain))
      BothMain <- unique(intersect(TargetMain,ReferenceMain))
      allCurrent <- unique(TargetMain,ReferenceMain)
      
      JustMyBackground <- unique(setdiff(TargetBackground,ReferenceBackground))
      JustCompetBackground <- unique(setdiff(ReferenceBackground,TargetBackground))
      BothBackground <- unique(intersect(TargetBackground,ReferenceBackground))
      allPrev <- unique(TargetBackground,ReferenceBackground)
      
      
      
      # Switched (from competitor)
      # Previous Month: Interactions on Pages owned by competitor
      # Current Month: Interactions on Pages owned only by our client only
      Switched <- intersect(JustMyMain,JustCompetBackground)
      
      # Won (from competitor)
      # Previous Month: Interactions on Pages owned by competitor
      # Current Month: Interactions on Pages owned only by our client only
      Won <- intersect(JustMyMain,BothBackground)
      
      # Loyal (purely to our client)
      # Previous Month: Interactions on Pages owned by our client only
      # Current Month: Interactions on Pages owned by our client only
      Loyal <- intersect(JustMyMain,JustMyBackground)
      
      
      # Acquired (“new new”)
      # Previous Month: No Interactions on any of the analyzed Pages
      # Current Month: Interactions on Pages owned by our client only
      Acquired <- setdiff(JustMyMain, unique(c(Switched,Won,Loyal)))
      
      UnLoyal <- intersect(BothMain,JustMyBackground)
      
      # Unstable
      # Previous Month: Interactions on Pages owned both by our client and competitor
      # Interactions on Pages owned both by our client and competitor
      Unstable <- intersect(BothMain,BothBackground)
      
      # Attracted (to our client)
      # Previous Month: Interactions on Pages only owned by competitor
      # Current Month: Interactions on Pages owned both by our client and competitor
      Attracted <- intersect(BothMain,JustCompetBackground)
      
      # True Intersect
      # No Interactions on any of the analyzed Pages
      # Interactions on Pages owned both by our client and competitor
      TrueIntersect <- setdiff(BothMain,unique(c(UnLoyal,Unstable,Attracted)))
      
      
      # Lost (to competition)
      # Interactions on Pages owned by our client
      # Interactions on Pages owned by competitor only
      Lost <- intersect(setdiff(ReferenceMain,TargetMain),setdiff(TargetBackground,ReferenceBackground))
      
      
      # Dead (completely)
      # 	Previous Month: Interactions on pages owned by our client
      # 	Current Month: No interactions on any of the analyzed pages
      Dead <- setdiff(TargetBackground, (c(ReferenceBackground,TargetMain,ReferenceMain)))
      
      
      
      AcquiredCount <- length(Acquired)
      SwitchedCount <- length(Switched)
      LoyalCount <- length(Loyal)
      WonCount <- length(Won)
      AttractedCount <- length(Attracted)
      UnstableCount <- length(Unstable)
      TrueIntersectCount <- length(TrueIntersect)
      UnLoyalCount <- length(UnLoyal)
      
      # TrueIntersectCount <- length(unique(TargetMain)) - (AcquiredCount + SwitchedCount + LoyalCount + UnstableCount + AttractedCount)
      
      LostCount <- length(Lost)
      DeadCount <- length(Dead)
      
      totalCount <-
        AcquiredCount +
        SwitchedCount +
        LoyalCount +
        WonCount +
        AttractedCount +
        UnstableCount  +
        TrueIntersectCount +
        UnLoyalCount
      
      
      
      overlapTmp <- data.frame(TargetName,
                               ReferenceName,
                               AcquiredCount,
                               AttractedCount,
                               SwitchedCount,
                               WonCount,
                               LoyalCount,
                               UnLoyalCount,
                               UnstableCount,
                               TrueIntersectCount,
                               LostCount,
                               DeadCount,
                               totalCount)
      
      overlap <- rbind(overlap, overlapTmp)
      
      ## Percentage Overlap
      
      overlapTmpPrct <- data.frame(TargetName,
                                   ReferenceName,
                                   round(100*(AcquiredCount/totalCount),2),
                                   round(100*(AttractedCount/totalCount),2),
                                   round(100*(SwitchedCount/totalCount),2),
                                   round(100*(WonCount/totalCount),2),
                                   round(100*(LoyalCount/totalCount),2),
                                   round(100*(UnLoyalCount/totalCount),2),
                                   round(100*(UnstableCount/totalCount),2),
                                   round(100*(TrueIntersectCount/totalCount),2),
                                   round(100*(totalCount/totalCount)))
      colnames(overlapTmpPrct) <- c(
        "TargetName",
        "ReferenceName",
        "AcquiredPct",
        "AttractedPct",
        "SwitchedPct",
        "WonPct",
        "LoyalPct",
        "UnLoyalPct",
        "UnstablePct",
        "TrueIntersectPct",
        "TotalaPct"
      )
      
      overlapPrct <- rbind(overlapPrct, overlapTmpPrct)
      colnames(overlapPrct) <- c(
        "TargetName",
        "ReferenceName",
        "AcquiredPct",
        "AttractedPct",
        "SwitchedPct",
        "WonPct",
        "LoyalPct",
        "UnLoyalPct",
        "UnstablePct",
        "TrueIntersectPct",
        "TotalaPct"
      )
      
      
      
      
      ## TopAtractedPost
      postsDetail<- data.frame()
      mainPageLikes <- subset(tblMainLikes, idPage == Target)
      
      for(postID in unique(mainPageLikes$post_id)) {
        likers <- subset(tblMainLikes, post_id == postID)
        likers <- likers$from_id
        likersLength <- length(likers)
        url <- paste0("https://www.facebook.com/", str_replace_all(postID, "_", "/posts/"))
        tblIntersectAcquired <- length(intersect(Acquired,likers))
        tblIntersectRatioAcquired <- tblIntersectAcquired/likersLength
        tblIntersectAcquiredPerct <- round(100*tblIntersectRatioAcquired,2)
        
        tblIntersectAttracted <- length(intersect(Attracted,likers))
        tblIntersectRatioAttracted <- tblIntersectAttracted/likersLength
        tblIntersectAttractedPerct <- round(100 *tblIntersectRatioAttracted,2)
        
        tblIntersectSwitched<- length(intersect(Switched,likers))
        tblIntersectRatioSwitched <- tblIntersectSwitched/likersLength
        tblIntersectSwitchedPerct <- round(100*tblIntersectRatioSwitched,2)
        
        tblIntersectWon<- length(intersect(Won,likers))
        tblIntersectRatioWon<- tblIntersectWon/likersLength
        tblIntersectWonPerct <- round(100*tblIntersectRatioWon,2)
        
        tblIntersectLoyal<- length(intersect(Loyal,likers))
        tblIntersectRatioLoyal<- tblIntersectLoyal/likersLength
        tblIntersectLoyalPerct <- round(100*tblIntersectRatioLoyal,2)
        
        tblIntersectUnLoyal<- length(intersect(UnLoyal,likers))
        tblIntersectRatioUnLoyal<- tblIntersectUnLoyal/likersLength
        tblIntersectUnLoyalPerct <- round(100*tblIntersectRatioUnLoyal,2)
        
        postsTmp <- data.frame(Target, TargetName,
                               Reference, ReferenceName, postID, url, likersLength,
                               tblIntersectAcquired, tblIntersectRatioAcquired,tblIntersectAcquiredPerct,
                               tblIntersectAttracted,tblIntersectRatioAttracted,tblIntersectAttractedPerct,
                               tblIntersectSwitched,tblIntersectRatioSwitched,tblIntersectSwitchedPerct,
                               tblIntersectWon,tblIntersectRatioWon,tblIntersectWonPerct,
                               tblIntersectLoyal,tblIntersectRatioLoyal,tblIntersectLoyalPerct,
                               tblIntersectUnLoyal,tblIntersectRatioUnLoyal,tblIntersectUnLoyalPerct
        )
        postsDetail <- rbind(postsDetail,postsTmp)
      }
      
      colnames(postsDetail) <- c("Target", "TargetName",
                                 "Reference", "ReferenceName", "id", "url", "likersLength",
                                 "tblIntersectAcquired", "tblIntersectRatioAcquired", "tblIntersectAcquiredPerct",
                                 "tblIntersectAttracted","tblIntersectRatioAttracted","tblIntersectAttractedPerct",
                                 "tblIntersectSwitched", "tblIntersectRatioSwitched", "tblIntersectSwitchedPerct",
                                 "tblIntersectWon", "tblIntersectRatioWon", "tblIntersectWonPerct",
                                 "tblIntersectLoyal", "tblIntersectRatioLoyal", "tblIntersectLoyalPerct",
                                 "tblIntersectUnLoyal", "tblIntersectRatioUnLoyal", "tblIntersectUnLoyalPerct"
      )
      
      postsDetail <- merge(postsDetail, tbl, by="id")
      postsDetail$link <- NULL
      postsDetail$totalInteraction <- postsDetail$comments_count + postsDetail$reactions_count + postsDetail$shares_count
      
      #	filename <- paste0(mainPage, "_to_",ReferenceName,"_postsDetail.csv")
      #	write.csv(postsDetail, filename)
      
      #filename <- paste0(TargetName, "_to_",ReferenceName,"_postsDetail.csv")
      #openxlsx::write.xlsx(postsDetail, filename, showNA = F, row.names = F)
      #write.csv(postsDetail, filename, row.names = F)
      postsDetail_all <- rbind(postsDetail_all, postsDetail)
      
    }
    
    overlap <- t(overlap)
    #	filename <- paste0(TargetName, "_overlap.csv")
    #	write.csv(overlap, filename)
    
    overlapPrct <- t(overlapPrct)
    #	filename <- paste0(TargetName, "_prct_overlap.csv")
    #	write.csv(overlapPrct, filename)
    
    filename <- paste0(TargetName, "_overlap.csv")
    #openxlsx::write.xlsx(overlap, filename, showNA = F, row.names = F)
    write.table(overlap, filename, sep = ";", col.names = T)
    
    filename <- paste0(TargetName, "_prct_overlap.csv")
    #openxlsx::write.xlsx(overlapPrct, filename,  showNA = F, row.names = F)
    write.table(overlapPrct, filename, sep = ";", col.names = T)
    
    postsDetail_sloucene <- rbind(postsDetail_sloucene, postsDetail_all)
  }
  postsDetail_sloucene$ReferenceName <- as.character(postsDetail_sloucene$ReferenceName)
  postsDetail_sloucene$TargetName <- as.character(postsDetail_sloucene$TargetName)
  postsDetail_sloucene$idPage <- as.character(postsDetail_sloucene$idPage)
  write.csv(postsDetail_sloucene, "_ALL_POSTS_Detail.csv", row.names = F)
  postsDetail_sloucene$message <- NULL
  write.xlsx(postsDetail_sloucene, "_ALL_POSTS_Detail.xlsx")

}