ratings_df <- read.csv(
"https://raw.githubusercontent.com/mj-nurse/mnurse-data_607/refs/heads/main/assignment-02a-movie-ratings/movie_ratings.csv",
na.strings = c("", "NULL")
)# View the first six rows
head(ratings_df)## respondent_label title rating
## 1 Person 1 The Drama 4
## 2 Person 1 Michael 3
## 3 Person 1 Obsession 5
## 4 Person 1 Backrooms 4
## 5 Person 1 The Odyssey 5
## 6 Person 1 Spider-Man: Brand New Day 3# Check the number of people, movies, and rows
length(unique(ratings_df$respondent_label))## [1] 22length(unique(ratings_df$title))## [1] 6nrow(ratings_df)## [1] 132# Check how many ratings are missing
sum(is.na(ratings_df$rating))## [1] 48## Mean Movie Rating
mean_movie_rating <- mean(ratings_df$rating, na.rm = TRUE)
mean_movie_rating## [1] 3.75library(dplyr)##
## Attaching package: 'dplyr'## The following objects are masked from 'package:stats':
##
## filter, lag## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, unionmovie_averages <- ratings_df %>%
group_by(title) %>%
summarise(
movie_avg = mean(rating, na.rm = TRUE),
.groups = "drop"
) %>%
mutate(
movie_avg_minus_mean_movie = movie_avg - mean_movie_rating
)
movie_averages## # A tibble: 6 × 3
## title movie_avg movie_avg_minus_mean_movie
## <chr> <dbl> <dbl>
## 1 Backrooms 3.1 -0.65
## 2 Michael 3.54 -0.212
## 3 Obsession 3.94 0.194
## 4 Spider-Man: Brand New Day 3.89 0.139
## 5 The Drama 3.58 -0.167
## 6 The Odyssey 4.15 0.404user_averages <- ratings_df %>%
group_by(respondent_label) %>%
summarise(
user_avg = mean(rating, na.rm = TRUE),
.groups = "drop"
) %>%
mutate(
user_avg_minus_mean_movie = user_avg - mean_movie_rating
)
user_averages <- user_averages %>%
arrange(as.numeric(sub("Person ", "", respondent_label)))
user_averages## # A tibble: 22 × 3
## respondent_label user_avg user_avg_minus_mean_movie
## <chr> <dbl> <dbl>
## 1 Person 1 4 0.25
## 2 Person 2 3.33 -0.417
## 3 Person 3 4.75 1
## 4 Person 4 4 0.25
## 5 Person 5 4.17 0.417
## 6 Person 6 3.5 -0.25
## 7 Person 7 4.67 0.917
## 8 Person 8 4 0.25
## 9 Person 9 2 -1.75
## 10 Person 10 4.8 1.05
## # ℹ 12 more rowspredicted_ratings <- ratings_df %>%
left_join(movie_averages, by = "title") %>%
left_join(user_averages, by = "respondent_label") %>%
filter(is.na(rating)) %>%
mutate(
global_baseline_estimate =
mean_movie_rating +
movie_avg_minus_mean_movie +
user_avg_minus_mean_movie
) %>%
arrange(
as.numeric(sub("Person ", "", respondent_label)),
desc(global_baseline_estimate)
)
predicted_ratings## respondent_label title rating movie_avg
## 1 Person 2 The Odyssey NA 4.153846
## 2 Person 2 The Drama NA 3.583333
## 3 Person 2 Backrooms NA 3.100000
## 4 Person 3 The Drama NA 3.583333
## 5 Person 3 Backrooms NA 3.100000
## 6 Person 4 The Odyssey NA 4.153846
## 7 Person 4 The Drama NA 3.583333
## 8 Person 4 Michael NA 3.538462
## 9 Person 4 Backrooms NA 3.100000
## 10 Person 6 Obsession NA 3.944444
## 11 Person 6 Michael NA 3.538462
## 12 Person 7 The Odyssey NA 4.153846
## 13 Person 7 Spider-Man: Brand New Day NA 3.888889
## 14 Person 7 Michael NA 3.538462
## 15 Person 8 Obsession NA 3.944444
## 16 Person 8 Spider-Man: Brand New Day NA 3.888889
## 17 Person 8 The Drama NA 3.583333
## 18 Person 8 Michael NA 3.538462
## 19 Person 8 Backrooms NA 3.100000
## 20 Person 9 The Drama NA 3.583333
## 21 Person 9 Michael NA 3.538462
## 22 Person 9 Backrooms NA 3.100000
## 23 Person 10 The Odyssey NA 4.153846
## 24 Person 11 The Drama NA 3.583333
## 25 Person 11 Backrooms NA 3.100000
## 26 Person 12 The Odyssey NA 4.153846
## 27 Person 12 Backrooms NA 3.100000
## 28 Person 14 The Odyssey NA 4.153846
## 29 Person 14 The Drama NA 3.583333
## 30 Person 14 Michael NA 3.538462
## 31 Person 14 Backrooms NA 3.100000
## 32 Person 15 The Odyssey NA 4.153846
## 33 Person 15 Obsession NA 3.944444
## 34 Person 15 Spider-Man: Brand New Day NA 3.888889
## 35 Person 15 Backrooms NA 3.100000
## 36 Person 16 The Drama NA 3.583333
## 37 Person 18 Michael NA 3.538462
## 38 Person 19 Obsession NA 3.944444
## 39 Person 19 The Drama NA 3.583333
## 40 Person 19 Michael NA 3.538462
## 41 Person 19 Backrooms NA 3.100000
## 42 Person 20 The Odyssey NA 4.153846
## 43 Person 20 Spider-Man: Brand New Day NA 3.888889
## 44 Person 20 The Drama NA 3.583333
## 45 Person 20 Backrooms NA 3.100000
## 46 Person 21 The Odyssey NA 4.153846
## 47 Person 21 Michael NA 3.538462
## 48 Person 21 Backrooms NA 3.100000
## movie_avg_minus_mean_movie user_avg user_avg_minus_mean_movie
## 1 0.4038462 3.333333 -0.4166667
## 2 -0.1666667 3.333333 -0.4166667
## 3 -0.6500000 3.333333 -0.4166667
## 4 -0.1666667 4.750000 1.0000000
## 5 -0.6500000 4.750000 1.0000000
## 6 0.4038462 4.000000 0.2500000
## 7 -0.1666667 4.000000 0.2500000
## 8 -0.2115385 4.000000 0.2500000
## 9 -0.6500000 4.000000 0.2500000
## 10 0.1944444 3.500000 -0.2500000
## 11 -0.2115385 3.500000 -0.2500000
## 12 0.4038462 4.666667 0.9166667
## 13 0.1388889 4.666667 0.9166667
## 14 -0.2115385 4.666667 0.9166667
## 15 0.1944444 4.000000 0.2500000
## 16 0.1388889 4.000000 0.2500000
## 17 -0.1666667 4.000000 0.2500000
## 18 -0.2115385 4.000000 0.2500000
## 19 -0.6500000 4.000000 0.2500000
## 20 -0.1666667 2.000000 -1.7500000
## 21 -0.2115385 2.000000 -1.7500000
## 22 -0.6500000 2.000000 -1.7500000
## 23 0.4038462 4.800000 1.0500000
## 24 -0.1666667 5.000000 1.2500000
## 25 -0.6500000 5.000000 1.2500000
## 26 0.4038462 4.000000 0.2500000
## 27 -0.6500000 4.000000 0.2500000
## 28 0.4038462 4.500000 0.7500000
## 29 -0.1666667 4.500000 0.7500000
## 30 -0.2115385 4.500000 0.7500000
## 31 -0.6500000 4.500000 0.7500000
## 32 0.4038462 4.500000 0.7500000
## 33 0.1944444 4.500000 0.7500000
## 34 0.1388889 4.500000 0.7500000
## 35 -0.6500000 4.500000 0.7500000
## 36 -0.1666667 3.200000 -0.5500000
## 37 -0.2115385 2.800000 -0.9500000
## 38 0.1944444 4.500000 0.7500000
## 39 -0.1666667 4.500000 0.7500000
## 40 -0.2115385 4.500000 0.7500000
## 41 -0.6500000 4.500000 0.7500000
## 42 0.4038462 3.500000 -0.2500000
## 43 0.1388889 3.500000 -0.2500000
## 44 -0.1666667 3.500000 -0.2500000
## 45 -0.6500000 3.500000 -0.2500000
## 46 0.4038462 3.333333 -0.4166667
## 47 -0.2115385 3.333333 -0.4166667
## 48 -0.6500000 3.333333 -0.4166667
## global_baseline_estimate
## 1 3.737179
## 2 3.166667
## 3 2.683333
## 4 4.583333
## 5 4.100000
## 6 4.403846
## 7 3.833333
## 8 3.788462
## 9 3.350000
## 10 3.694444
## 11 3.288462
## 12 5.070513
## 13 4.805556
## 14 4.455128
## 15 4.194444
## 16 4.138889
## 17 3.833333
## 18 3.788462
## 19 3.350000
## 20 1.833333
## 21 1.788462
## 22 1.350000
## 23 5.203846
## 24 4.833333
## 25 4.350000
## 26 4.403846
## 27 3.350000
## 28 4.903846
## 29 4.333333
## 30 4.288462
## 31 3.850000
## 32 4.903846
## 33 4.694444
## 34 4.638889
## 35 3.850000
## 36 3.033333
## 37 2.588462
## 38 4.694444
## 39 4.333333
## 40 4.288462
## 41 3.850000
## 42 3.903846
## 43 3.638889
## 44 3.333333
## 45 2.850000
## 46 3.737179
## 47 3.121795
## 48 2.683333recommendations <- predicted_ratings %>%
group_by(respondent_label) %>%
slice_max(
order_by = global_baseline_estimate,
n = 1,
with_ties = TRUE
) %>%
ungroup() %>%
arrange(as.numeric(sub("Person ", "", respondent_label))) %>%
select(
respondent_label,
title,
global_baseline_estimate
)
recommendations## # A tibble: 17 × 3
## respondent_label title global_baseline_estimate
## <chr> <chr> <dbl>
## 1 Person 2 The Odyssey 3.74
## 2 Person 3 The Drama 4.58
## 3 Person 4 The Odyssey 4.40
## 4 Person 6 Obsession 3.69
## 5 Person 7 The Odyssey 5.07
## 6 Person 8 Obsession 4.19
## 7 Person 9 The Drama 1.83
## 8 Person 10 The Odyssey 5.20
## 9 Person 11 The Drama 4.83
## 10 Person 12 The Odyssey 4.40
## 11 Person 14 The Odyssey 4.90
## 12 Person 15 The Odyssey 4.90
## 13 Person 16 The Drama 3.03
## 14 Person 18 Michael 2.59
## 15 Person 19 Obsession 4.69
## 16 Person 20 The Odyssey 3.90
## 17 Person 21 The Odyssey 3.74The Global Baseline Estimate combines the Mean Movie Rating of 3.75 with each movie’s and each user’s rating relative to that average. I calculated estimates for the 48 missing ratings I had and selected the highest estimate for each person. This produced recommendations for the 17 respondents who had not seen 1 of the 6 films. The other 5 respondents had rated all 6 movies. In conclusion, The Odyssey was recommended most frequently, appearing in nine recommendations.