Top 3 movies by Occupation

#merge the dataset

UData_User<-merge(U.Data,y = U.User,by = "User_ID")

# group by occupation,item_id 
groupby<-by(UData_User,list(UData_User$occupation,UData_User$Item_ID),FUN = function(x) 
  {
  data.frame(occupation=unique(x$occupation),
  Item_ID=unique(x$Item_ID),
  mean_rating=mean(x$Rating),
  Count_Item_ID=nrow(x)
  )
  })

Top3Occupation<-do.call(rbind,groupby)
Top3Occupation<-Top3Occupation[with(Top3Occupation,order(occupation,-mean_rating,-Count_Item_ID)),]


# pick top 3 

Top3Occupation<-by(Top3Occupation,list(Top3Occupation$occupation),head,n=3) 
Top3Occupation<-do.call(rbind.data.frame,Top3Occupation)
rownames(Top3Occupation)<-NULL

# Get the names of the movie

Top3Occupation<-merge(Top3Occupation,U.Item[,1:2],by.x = "Item_ID",by.y = "movie_id",sort = FALSE)
Top3Occupation<-Top3Occupation[with(Top3Occupation,order(occupation,-mean_rating,-Count_Item_ID)),]
knitr::kable(Top3Occupation,caption = "Top 3 Movies by Occupation")

write.csv(x = Top3Occupation,file = "Top3Occupation.csv")

Top 3 movies by Genre

#reversing one-hot encoding to match to the dataset

rev_one_hot<-as.data.frame(which(U.Item[,5:23]==1,arr.ind = T))
rev_one_hot$genre_transformed<-names(U.Item[,5:23])[rev_one_hot[order(rev_one_hot[,1]),2]]
Genre<-merge(U.Item,rev_one_hot,by.x = "movie_id",by.y = "row")

Selecting relevant columns from Genre

# merging Genre with UData on Item Id 
UGenre<-Genre[,c(1,2,25)]
UGenreUser<-merge(UGenre,U.Data,by.x ="movie_id",by.y = "Item_ID" )

#group by Genre

groupby<-by(UGenreUser,list(UGenreUser$genre_transformed,UGenreUser$movie_id),FUN = function(x) 
  {
  data.frame(genre_transformed=unique(x$genre_transformed),
  Item_ID=unique(x$movie_id),
  movie_title=unique(x$movie_title),
  mean_rating=mean(x$Rating),
  Count_Item_ID=nrow(x)
  )
  })

Top3Genre<-do.call(rbind,groupby)
Top3Genre<-Top3Genre[with(Top3Genre,order(genre_transformed,-mean_rating,-Count_Item_ID)),]


# pick top 3 

Top3Genre<-by(Top3Genre,list(Top3Genre$genre_transformed),head,n=3) 
Top3Genre<-do.call(rbind.data.frame,Top3Genre)
rownames(Top3Genre)<-NULL

knitr::kable(Top3Genre,caption = "Top 3 Movies by Genre")

write.csv(x = Top3Genre,file = "Top3Genre.csv")

Top 3 movies by Occupation,Genre

Dataset created in Top 3 Movies by Genre is reused

# merging data 

UGenreUser_Occupation<-merge(UGenreUser,U.User,by="User_ID")

#  group by occupation,genre

groupby<-by(UGenreUser_Occupation,list(UGenreUser_Occupation$occupation,UGenreUser_Occupation$genre_transformed,UGenreUser_Occupation$movie_id),FUN = function(x) 
  {
  data.frame(occupation=unique(x$occupation),
  genre_transformed=unique(x$genre_transformed),
  movie_id=unique(x$movie_id),
  mean_rating=mean(x$Rating),
  Count_Item_ID=nrow(x)
  )
  })

Top3OccupationGenre<-do.call(rbind,groupby)
Top3OccupationGenre<-Top3OccupationGenre[with(Top3OccupationGenre,order(occupation,genre_transformed,-mean_rating,-Count_Item_ID)),]

Top3OccupationGenre<-aggregate(Top3OccupationGenre,by=list(Top3OccupationGenre$occupation,Top3OccupationGenre$genre_transformed),FUN = head,n=3)

kable(Top3OccupationGenre,caption = "Top 3 Movies by Occupation, Genre")

For occupation and genre data has been formatted so that the top 3 movies are shown with their items ids.

Top 3 Movies by Age Group

# find the oldest user

max_age=max(U.User$age)

U.User$age_bracket<- cut(U.User$age, breaks = c(0,6, 12, 18, 30, 50,(max_age+1)),
      labels = c("0-6", "6-12", "12-18", "18-30","30-50","50+"),
      right = T)

# grouping by Age bracket


UAgeUser<-merge(U.User,U.Data,by.x ="User_ID",by.y = "User_ID" )
groupby<-by(UAgeUser,list(UAgeUser$age_bracket,UAgeUser$Item_ID),FUN = function(x) 
  {
  data.frame(age_bracket=unique(x$age_bracket),
  Item_ID=unique(x$Item_ID),
  mean_rating=mean(x$Rating),
  Count_Item_ID=nrow(x)
  )
  })

Top3Age_Group<-do.call(rbind,groupby)
  Top3Age_Group<-Top3Age_Group[with(Top3Age_Group,order(age_bracket,-mean_rating,-Count_Item_ID)),]


# pick top 3 

Top3Age_Group<-by(Top3Age_Group,list(Top3Age_Group$age_bracket),head,n=3) 
Top3Age_Group<-do.call(rbind.data.frame,Top3Age_Group)
rownames(Top3Age_Group)<-NULL


# Get the names of the movie

Top3Age_Group<-merge(Top3Age_Group,U.Item[,1:2],by.x = "Item_ID",by.y = "movie_id",sort = FALSE)
Top3Age_Group<-Top3Age_Group[with(Top3Age_Group,order(age_bracket,-mean_rating,-Count_Item_ID)),]


knitr::kable(Top3Age_Group,caption = "Top 3 Movies by Age group")

write.csv(x = Top3Age_Group,file = "Top3Age_Group.csv")

Top 3 Genres released in Summer

Using the release date column to find movies released in Summer

#subsetting the data frame by selecting movies released in Summer

Genre$release_date<-as.Date(Genre$release_date,format="%d-%b-%y")
Genre$release_date<-format (Genre$release_date, "%b")
Genre_Summer<-Genre[Genre$release_date %in% c("May","June","July"),]

Genre_Summer<-Genre_Summer[,c(1:3,25)]

# merging data 

Genre_Summer<-merge(Genre_Summer,U.Data,by.x = "movie_id","Item_ID")


groupby<-by(Genre_Summer,list(Genre_Summer$genre_transformed),FUN = function(x) 
  {
  data.frame(genre_transformed=unique(x$genre_transformed),
  mean_rating=mean(x$Rating),
  Count_Item_ID=nrow(x)
  )
  })

Top3Genre_Summer<-do.call(rbind,groupby)
Top3Genre_Summer<-Top3Genre_Summer[with(Top3Genre_Summer,order(-mean_rating,-Count_Item_ID)),]


# pick top 3 

Top3Genre_Summer<-head(Top3Genre_Summer,n = 3) 
rownames(Top3Genre_Summer)<-NULL

knitr::kable(Top3Genre_Summer,caption = "Top 3 Genres released in Summer")

write.csv(x = Top3Genre_Summer,file = "Top3Genre_Summer.csv")

For each genre, co-occuring (top2) genre

# correlation matrix would give us what genres are closed to what genres 
Correlation_Matrix<-cor(U.Item[,5:23])
Correlation_Matrix<-as.data.frame(Correlation_Matrix)

find_co_occuring<-function(cname)
{
  top2<-tail(head(with(Correlation_Matrix,order(-cname)),n=3),n = 2)
  rownames(Correlation_Matrix)[top2]
}

Top2CoOccuring<-apply(Correlation_Matrix, 2, find_co_occuring)

knitr::kable(Top2CoOccuring,caption = "Top 2 Co-Occuring Genres for each genre")

write.csv(x = Top2CoOccuring,file = "Top2CoOccuring.csv")

This means that whenever the movie has been rated for action it has also been rated for Adventure and Sci-Fi