Julia Ferris
2023-09-18
Two packages are needed. readr is used to import the data. gt is used to format the data nicely.
movies <- read_csv("https://raw.githubusercontent.com/juliaDataScience-22/cuny-fall-23/manage-acquire-data/newMovieRatings.csv",
show_col_types = FALSE)
gt(movies) |>
tab_header(
title = "Table 1",
subtitle = "Movie Ratings"
)| Table 1 | ||||||
| Movie Ratings | ||||||
| Critic | AvatarWater | TopGunMaverick | Oppenheimer | SoundOfFreedom | Barbie | Boogeyman |
|---|---|---|---|---|---|---|
| Jan | 3 | NA | 4 | 4 | 2 | NA |
| Jill | 4 | 5 | NA | 5 | 4 | 2 |
| Claire | NA | 3 | NA | 5 | 4 | 5 |
| Dan | 5 | 5 | 3 | 5 | 3 | NA |
| Sean | 5 | 4 | 4 | NA | NA | 4 |
Note that the echo = FALSE parameter was added to the
code chunk to prevent printing of the R code that generated the
plot.
movieAvgs <- data.frame(movie = c("AvatarWater","TopGunMaverick",
"Oppenheimer","SoundOfFreedom",
"Barbie","Boogeyman"),
average = c(mean(movies$AvatarWater, na.rm = TRUE),
mean(movies$TopGunMaverick, na.rm = TRUE),
mean(movies$Oppenheimer, na.rm = TRUE),
mean(movies$SoundOfFreedom, na.rm = TRUE),
mean(movies$Barbie, na.rm = TRUE),
mean(movies$Boogeyman, na.rm = TRUE)))
userAvgs <- data.frame(user = c("Jan","Jill","Claire","Dan","Sean"),
average = c(mean(c(movies$AvatarWater[1],
movies$TopGunMaverick[1],
movies$Oppenheimer[1],
movies$SoundOfFreedom[1],
movies$Barbie[1],
movies$Boogeyman[1]),
na.rm = TRUE),
mean(c(movies$AvatarWater[2],
movies$TopGunMaverick[2],
movies$Oppenheimer[2],
movies$SoundOfFreedom[2],
movies$Barbie[2],
movies$Boogeyman[2]),
na.rm = TRUE),
mean(c(movies$AvatarWater[3],
movies$TopGunMaverick[3],
movies$Oppenheimer[3],
movies$SoundOfFreedom[3],
movies$Barbie[3],
movies$Boogeyman[3]),
na.rm = TRUE),
mean(c(movies$AvatarWater[4],
movies$TopGunMaverick[4],
movies$Oppenheimer[4],
movies$SoundOfFreedom[4],
movies$Barbie[4],
movies$Boogeyman[4]),
na.rm = TRUE),
mean(c(movies$AvatarWater[5],
movies$TopGunMaverick[5],
movies$Oppenheimer[5],
movies$SoundOfFreedom[5],
movies$Barbie[5],
movies$Boogeyman[5]),
na.rm = TRUE)))
meanMovieRating <- sum(c(movies$AvatarWater,
movies$TopGunMaverick,
movies$Oppenheimer,
movies$SoundOfFreedom,
movies$Barbie,
movies$Boogeyman),na.rm=TRUE) / sum(!is.na(movies[2:7]))
movieAvgs$movieAvgMinusMeanMovie <- c(movieAvgs$average[1] - meanMovieRating,
movieAvgs$average[2] - meanMovieRating,
movieAvgs$average[3] - meanMovieRating,
movieAvgs$average[4] - meanMovieRating,
movieAvgs$average[5] - meanMovieRating,
movieAvgs$average[6] - meanMovieRating)
userAvgs$userAvgMinusMeanMovie <- c(userAvgs$average[1] - meanMovieRating,
userAvgs$average[2] - meanMovieRating,
userAvgs$average[3] - meanMovieRating,
userAvgs$average[4] - meanMovieRating,
userAvgs$average[5] - meanMovieRating)
gt(movieAvgs) |>
tab_header(
title = "Table 2",
subtitle = "Movie Averages"
)| Table 2 | ||
| Movie Averages | ||
| movie | average | movieAvgMinusMeanMovie |
|---|---|---|
| AvatarWater | 4.250000 | 0.2500000 |
| TopGunMaverick | 4.250000 | 0.2500000 |
| Oppenheimer | 3.666667 | -0.3333333 |
| SoundOfFreedom | 4.750000 | 0.7500000 |
| Barbie | 3.250000 | -0.7500000 |
| Boogeyman | 3.666667 | -0.3333333 |
| Table 3 | ||
| User Averages | ||
| user | average | userAvgMinusMeanMovie |
|---|---|---|
| Jan | 3.25 | -0.75 |
| Jill | 4.00 | 0.00 |
| Claire | 4.25 | 0.25 |
| Dan | 4.20 | 0.20 |
| Sean | 4.25 | 0.25 |
The function can accomplish two things. The first is to rate a movie that a person has not seen. The second is to recommend a movie of the ones a person has not seen. The function is based on the algorithm known as the Global Baseline Estimate.
Please note: This algorithm will only work with the data set used. If another data set it used, it will not work. However, this function can be modified for other data sets. Also, if multiple movies tie for the same rating, only one of them will be recommended to the person. Lastly, error checking is not included in the function because the function is tested with pre-determined input.
ratingAlgorithm <- function(name, determine) {
if(determine == "rate")
{
personName <- name
personName <- switch(personName,
"Jan"="1",
"Jill"="2",
"Claire"="3",
"Dan"="4",
"Sean"="5",
"other")
personName <- suppressWarnings(as.numeric(personName))
movieTitle <- 0
ifelse(is.na(movies$AvatarWater[personName]), movieTitle <- 1, movieTitle <- movieTitle)
ifelse(is.na(movies$TopGunMaverick[personName]), movieTitle <- 2, movieTitle <- movieTitle)
ifelse(is.na(movies$Oppenheimer[personName]), movieTitle <- 3, movieTitle <- movieTitle)
ifelse(is.na(movies$SoundOfFreedom[personName]), movieTitle <- 4, movieTitle <- movieTitle)
ifelse(is.na(movies$Barbie[personName]), movieTitle <- 5, movieTitle <- movieTitle)
ifelse(is.na(movies$Boogeyman[personName]), movieTitle <- 6, movieTitle <- movieTitle)
movieName <- as.character(movieTitle)
movieName <- switch(movieName,
"1"="Avatar: The Way of Water",
"2"="Top Gun: Maverick",
"3"="Oppenheimer",
"4"="Sound of Freedom",
"5"="Barbie",
"6"="The Boogeyman",
"other")
rating <- meanMovieRating +
movieAvgs$movieAvgMinusMeanMovie[movieTitle] +
userAvgs$userAvgMinusMeanMovie[personName]
paste(name, "would give", movieName, "a rating of", round(rating, digits = 2))
}else if(determine == "recommend")
{
personName <- name
personName <- switch(personName,
"Jan"="1",
"Jill"="2",
"Claire"="3",
"Dan"="4",
"Sean"="5",
"other")
personName <- suppressWarnings(as.numeric(personName))
check1 <- 0
check2 <- 0
check3 <- 0
check4 <- 0
check5 <- 0
check6 <- 0
ifelse(is.na(movies$AvatarWater[personName]), check1 <- 1, check1 <- 0)
ifelse(is.na(movies$TopGunMaverick[personName]), check2 <- 2, check2 <- 0)
ifelse(is.na(movies$Oppenheimer[personName]), check3 <- 3, check3 <- 0)
ifelse(is.na(movies$SoundOfFreedom[personName]), check4 <- 4, check4 <- 0)
ifelse(is.na(movies$Barbie[personName]), check5 <- 5, check5 <- 0)
ifelse(is.na(movies$Boogeyman[personName]), check6 <- 6, check6 <- 0)
checks <- c(check1, check2, check3, check4, check5, check6)
ratings <- c()
movieNames <- c()
for (check in checks){
if(check != 0)
{
movieTitle <- check
rating <- meanMovieRating +
movieAvgs$movieAvgMinusMeanMovie[movieTitle] +
userAvgs$userAvgMinusMeanMovie[personName]
movieName <- as.character(movieTitle)
movieName <- switch(movieName,
"1"="Avatar: The Way of Water",
"2"="Top Gun: Maverick",
"3"="Oppenheimer",
"4"="Sound of Freedom",
"5"="Barbie",
"6"="The Boogeyman",
"other")
ratings <- c(ratings, rating)
movieNames <- c(movieNames, movieName)
}
}
compareMovies <- data.frame(movieRatings = c(ratings),
theMovieNames = c(movieNames))
bestMovie1 <- max(compareMovies$movieRatings,is.na = TRUE)
count = 1
bestMovie2 <- 0
for (rating in compareMovies$movieRatings)
{
if(rating == bestMovie1)
{
bestMovie2 <- compareMovies$theMovieNames[count]
break
}
count <- count + 1
}
paste(name, "should watch", bestMovie2)
}
}Each person had missing values. People who missed two or more movies will be given a recommendation. People who only missed on movie will have a rating created for the one they missed.
## [1] "Jan should watch Top Gun: Maverick"## [1] "Jill would give Oppenheimer a rating of 3.67"## [1] "Claire should watch Avatar: The Way of Water"## [1] "Dan would give The Boogeyman a rating of 3.87"## [1] "Sean should watch Sound of Freedom"Here are a few ways to add to this information:
Lander, Jared P. R for Everyone: Advanced Analytics and Graphics. Addison Wesley, 2017.