Data Dive Week 4
2025-02-07
data <-read.csv("movies_metadata.csv")
# Replace "[]" with NA
data[data == "[]"] <- NA
library(tidyverse)## Warning: package 'tidyverse' was built under R version 4.3.3## Warning: package 'ggplot2' was built under R version 4.3.3## Warning: package 'tibble' was built under R version 4.3.3## Warning: package 'tidyr' was built under R version 4.3.3## Warning: package 'readr' was built under R version 4.3.3## Warning: package 'purrr' was built under R version 4.3.3## Warning: package 'dplyr' was built under R version 4.3.3## Warning: package 'stringr' was built under R version 4.3.3## Warning: package 'forcats' was built under R version 4.3.3## Warning: package 'lubridate' was built under R version 4.3.3## ── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
## ✔ dplyr 1.1.4 ✔ readr 2.1.5
## ✔ forcats 1.0.0 ✔ stringr 1.5.1
## ✔ ggplot2 3.5.1 ✔ tibble 3.2.1
## ✔ lubridate 1.9.4 ✔ tidyr 1.3.1
## ✔ purrr 1.0.2
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag() masks stats::lag()
## ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errorslibrary(dplyr)
library(ggplot2)Taking Random Samples
sample_size <- floor(0.5 * nrow(data))
df_1 <- data.frame(data[sample(nrow(data), sample_size, replace = TRUE), ])
df_2 <- data.frame(data[sample(nrow(data), sample_size, replace = TRUE), ])
df_3 <- data.frame(data[sample(nrow(data), sample_size, replace = TRUE), ])
df_4 <- data.frame(data[sample(nrow(data), sample_size, replace = TRUE), ])
df_5 <- data.frame(data[sample(nrow(data), sample_size, replace = TRUE), ])Summary of Occupation to Budgets Across Sub Samples
# For df_1
df_1 %>%
mutate(budget = as.numeric(budget)) %>%
filter(!is.na(budget)) %>% # Remove NAs before calculating summary
summarise(
Min = min(budget),
Max = max(budget),
Mean = mean(budget),
Median = median(budget)
)## Min Max Mean Median
## 1 0 3.8e+08 4375417 0# For df_2
df_2 %>%
mutate(budget = as.numeric(budget)) %>%
filter(!is.na(budget)) %>%
summarise(
Min = min(budget),
Max = max(budget),
Mean = mean(budget),
Median = median(budget)
)## Warning: There was 1 warning in `mutate()`.
## ℹ In argument: `budget = as.numeric(budget)`.
## Caused by warning:
## ! NAs introduced by coercion## Min Max Mean Median
## 1 0 3.8e+08 4241655 0# For df_3
df_3 %>%
mutate(budget = as.numeric(budget)) %>%
filter(!is.na(budget)) %>%
summarise(
Min = min(budget),
Max = max(budget),
Mean = mean(budget),
Median = median(budget)
)## Warning: There was 1 warning in `mutate()`.
## ℹ In argument: `budget = as.numeric(budget)`.
## Caused by warning:
## ! NAs introduced by coercion## Min Max Mean Median
## 1 0 3.8e+08 4316634 0# For df_4
df_4 %>%
mutate(budget = as.numeric(budget)) %>%
filter(!is.na(budget)) %>%
summarise(
Min = min(budget),
Max = max(budget),
Mean = mean(budget),
Median = median(budget)
)## Warning: There was 1 warning in `mutate()`.
## ℹ In argument: `budget = as.numeric(budget)`.
## Caused by warning:
## ! NAs introduced by coercion## Min Max Mean Median
## 1 0 3.8e+08 4254095 0# For df_5
df_5 %>%
mutate(budget = as.numeric(budget)) %>%
filter(!is.na(budget)) %>%
summarise(
Min = min(budget),
Max = max(budget),
Mean = mean(budget),
Median = median(budget)
)## Warning: There was 1 warning in `mutate()`.
## ℹ In argument: `budget = as.numeric(budget)`.
## Caused by warning:
## ! NAs introduced by coercion## Min Max Mean Median
## 1 0 3.8e+08 4300509 0
df_1$budget <- as.numeric(df_1$budget)
df_2$budget <- as.numeric(df_2$budget)## Warning: NAs introduced by coerciondf_3$budget <- as.numeric(df_3$budget)## Warning: NAs introduced by coerciondf_4$budget <- as.numeric(df_4$budget)## Warning: NAs introduced by coerciondf_5$budget <- as.numeric(df_5$budget)## Warning: NAs introduced by coerciondf_1 <- df_1[!is.na(df_1$budget), ]
df_2 <- df_2[!is.na(df_2$budget), ]
df_3 <- df_3[!is.na(df_3$budget), ]
df_4 <- df_4[!is.na(df_4$budget), ]
df_5 <- df_5[!is.na(df_5$budget), ]
boxplot(df_1$budget, main = "Budget Distribution in df_1", ylab = "Budget")
boxplot(df_2$budget, main = "Budget Distribution in df_2", ylab = "Budget")
boxplot(df_3$budget, main = "Budget Distribution in df_3", ylab = "Budget")
boxplot(df_4$budget, main = "Budget Distribution in df_4", ylab = "Budget")
boxplot(df_5$budget, main = "Budget Distribution in df_5", ylab = "Budget")
df_1 |>
ggplot() +
geom_histogram(mapping = aes(x = budget),
colour = 'white') +
scale_x_continuous(labels = scales::dollar_format()) + # Use dollar_format without dividing by 1000
labs(title = "Budgets of Movies in df_1") +
theme(plot.subtitle = element_text(color = 'darkgray'))## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
df_2 |>
ggplot() +
geom_histogram(mapping = aes(x = budget),
colour = 'white') +
scale_x_continuous(labels = scales::dollar_format()) + # Use dollar_format without dividing by 1000
labs(title = "Budgets of Movies in df_2") +
theme(plot.subtitle = element_text(color = 'darkgray'))## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
df_3 |>
ggplot() +
geom_histogram(mapping = aes(x = budget),
colour = 'white') +
scale_x_continuous(labels = scales::dollar_format()) + # Use dollar_format without dividing by 1000
labs(title = "Budgets of Movies in df_3") +
theme(plot.subtitle = element_text(color = 'darkgray'))## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
df_4 |>
ggplot() +
geom_histogram(mapping = aes(x = budget),
colour = 'white') +
scale_x_continuous(labels = scales::dollar_format()) + # Use dollar_format without dividing by 1000
labs(title = "Budgets of Movies in df_4") +
theme(plot.subtitle = element_text(color = 'darkgray'))## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
df_5 |>
ggplot() +
geom_histogram(mapping = aes(x = budget),
colour = 'white') +
scale_x_continuous(labels = scales::dollar_format()) + # Use dollar_format without dividing by 1000
labs(title = "Budgets of Movies in df_5") +
theme(plot.subtitle = element_text(color = 'darkgray'))## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
Insights
The data suggests that the majority of movies in each sub sample has a $0 budget, indicating that they may be indie movies or fan films.
When looking for anomalies in the budget, I found that the averages are very consistent across all samples. One anomaly that can be seen though is that som sub samples have a maximum budget of $380,000,000 while the rest of the sub samples have a maximum budget of $270,000,000 or $300,000,000.
The $380,000,000 budgets are also outliers in their respective data samples indicating that they are very rare budgets.
df_1$runtime <- as.numeric(df_1$runtime)
df_1$popularity <- as.numeric(df_1$popularity)
df_1 <- df_1[!is.na(df_1$runtime) & !is.na(df_1$popularity), ]
ggplot(df_1, aes(x = runtime, y = popularity)) +
geom_point(color = "#69b3a2", size = 3, alpha = 0.7) +
ggtitle("Movie Runtime vs Popularity Rating") +
xlab("Movie Runtime (minutes)") +
ylab("Popularity Rating") +
theme_minimal() +
scale_x_continuous(breaks = seq(0, max(df_1$runtime, na.rm = TRUE), by = 50)) +
scale_y_continuous(breaks = seq(0, max(df_1$popularity, na.rm = TRUE), by = 50))
df_2$runtime <- as.numeric(df_2$runtime)
df_2$popularity <- as.numeric(df_2$popularity)
df_2 <- df_2[!is.na(df_2$runtime) & !is.na(df_2$popularity), ]
ggplot(df_2, aes(x = runtime, y = popularity)) +
geom_point(color = "#69b3a2", size = 3, alpha = 0.7) +
ggtitle("Movie Runtime vs Popularity Rating") +
xlab("Movie Runtime (minutes)") +
ylab("Popularity Rating") +
theme_minimal() +
scale_x_continuous(breaks = seq(0, max(df_2$runtime, na.rm = TRUE), by = 50)) +
scale_y_continuous(breaks = seq(0, max(df_2$popularity, na.rm = TRUE), by = 50))
df_3$runtime <- as.numeric(df_3$runtime)
df_3$popularity <- as.numeric(df_3$popularity)
df_3 <- df_3[!is.na(df_3$runtime) & !is.na(df_3$popularity), ]
ggplot(df_3, aes(x = runtime, y = popularity)) +
geom_point(color = "#69b3a2", size = 3, alpha = 0.7) +
ggtitle("Movie Runtime vs Popularity Rating") +
xlab("Movie Runtime (minutes)") +
ylab("Popularity Rating") +
theme_minimal() +
scale_x_continuous(breaks = seq(0, max(df_3$runtime, na.rm = TRUE), by = 50)) +
scale_y_continuous(breaks = seq(0, max(df_3$popularity, na.rm = TRUE), by = 50))
df_4$runtime <- as.numeric(df_4$runtime)
df_4$popularity <- as.numeric(df_4$popularity)
df_4 <- df_4[!is.na(df_4$runtime) & !is.na(df_4$popularity), ]
ggplot(df_4, aes(x = runtime, y = popularity)) +
geom_point(color = "#69b3a2", size = 3, alpha = 0.7) +
ggtitle("Movie Runtime vs Popularity Rating") +
xlab("Movie Runtime (minutes)") +
ylab("Popularity Rating") +
theme_minimal() +
scale_x_continuous(breaks = seq(0, max(df_4$runtime, na.rm = TRUE), by = 50)) +
scale_y_continuous(breaks = seq(0, max(df_4$popularity, na.rm = TRUE), by = 50))
df_5$runtime <- as.numeric(df_5$runtime)
df_5$popularity <- as.numeric(df_5$popularity)
df_5 <- df_5[!is.na(df_5$runtime) & !is.na(df_5$popularity), ]
ggplot(df_5, aes(x = runtime, y = popularity)) +
geom_point(color = "#69b3a2", size = 3, alpha = 0.7) +
ggtitle("Movie Runtime vs Popularity Rating") +
xlab("Movie Runtime (minutes)") +
ylab("Popularity Rating") +
theme_minimal() +
scale_x_continuous(breaks = seq(0, max(df_5$runtime, na.rm = TRUE), by = 50)) +
scale_y_continuous(breaks = seq(0, max(df_5$popularity, na.rm = TRUE), by = 50))
Insights:
The dataset is skewed towards movies with runtimes of around 100-150 minutes.
There are some data points that extend the average runtime but have little popularity which we can call outliers.
The majority of the data points are clustered in around the same area across all subsamples when we compare the Runtime with Popularity ratings.
How Might this Investigation Affect Drawing Conclusions
- In future analyses, I’ll separate outliers (like extreme runtimes or budgets) to avoid skewing conclusions, as they may represent rarer cases.
I’ll verify unusual values (like $0 budgets) to ensure they’re accurate and not data entry errors, especially when drawing conclusions about indie or fan films.
I’ll segment the data by genre or production company to better understand how different factors (runtime, budget) influence movie success across various categories.