Complete Data Project in R (Example) | Import, Manipulation, Graphics, Statistical Analysis & PCA

How to conduct a complete data project in the R programming language. Overview of the courses offered by Statistics Globe: https://statisticsglobe.com/courses

Link to data: https://www.kaggle.com/datasets/nelgi...

R Code:

install.packages("dplyr") # Install & load dplyr
library(dplyr)

install.packages("ggplot2") # Install & load ggplot2
library(ggplot2)

my_path ＜- "C:/Users/Joachim Schork/Dropbox/Jock/Data Sets/" # Path

my_spotify ＜- read.csv(paste0(my_path, # Import data
"Most_Streamed_Spotify_Songs_2024.csv"))

my_spotify ＜- my_spotify %＞% # Modify data
select("Spotify.Streams", "Spotify.Playlist.Count", # Select relevant columns
"Spotify.Playlist.Reach", "YouTube.Views",
"YouTube.Likes", "TikTok.Posts", "TikTok.Likes",
"TikTok.Views", "YouTube.Playlist.Reach",
"AirPlay.Spins", "Deezer.Playlist.Reach",
"Pandora.Streams", "Shazam.Counts",
"Explicit.Track") %＞%
select(sort(names(.))) %＞% # Sort columns
mutate(across(where(is.character), ~na_if(.x, ""))) %＞% # Replace empty cells with NA
mutate(across(where(is.character), # Modify commas & class
~as.numeric(gsub(",", "", .x)))) %＞%
mutate(across(-Explicit.Track, ~ifelse(. ＜ 10, NA, .))) %＞% # Replace below 10 with NA
na.omit() # Remove rows with NA

ggplot(my_spotify, # Draw density plot of Streams
aes(x = Spotify.Streams)) +
geom_density()

ggplot(my_spotify, # Grouped densities
aes(x = Spotify.Streams,
col = factor(Explicit.Track))) +
geom_density()

my_spotify %＞% # Improve layout of plot
mutate(Explicit.Track = ifelse(Explicit.Track == 1, # Change legend items
"Yes", "No")) %＞%
ggplot(aes(x = Spotify.Streams, # Specify plot variables
color = Explicit.Track)) +
geom_density(linewidth = 1) + # Draw density
labs(title = "Density Plot of Spotify Streams by Explicit Track", # Plot labels
x = "Spotify Streams",
y = "Density",
color = "") +
theme_minimal(base_size = 14) + # Specify plot theme
theme(plot.title = element_text(face = "bold", hjust = 0.5),
legend.position = "top")

my_mod ＜- lm(Spotify.Streams ~ ., my_spotify) # Estimate linear regression
summary(my_mod) # Print summary statistics

my_spotify_x ＜- my_spotify %＞% # Create subset of predictors
select(-Spotify.Streams)

pca_result ＜- prcomp(my_spotify_x, scale. = TRUE) # Apply PCA
summary(pca_result) # Summary of PCA results

my_spotify_pca ＜- data.frame(Spotify.Streams = my_spotify$Spotify.Streams, # Combine data
pca_result$x[ , 1:5])

my_mod_pca ＜- lm(Spotify.Streams ~ ., data = my_spotify_pca) # Apply PCR
summary(my_mod_pca) # Summary of PCR results

Follow me on Social Media:
LinkedIn – Joachim Schork Profile:   / linkedin   – Statistics Globe Page:   / statisticsglobe  
LinkedIn – R Programming Group for Discussions & Questions:   / 12555223  
LinkedIn – Python Programming Group for Discussions & Questions:   / 12673534  
X (Formerly Twitter): https://x.com/JoachimSchork
Facebook – Joachim Schork Profile:   / facebook   – Statistics Globe Page:   / statisticsglobecom  
Facebook – R Programming Group for Discussions & Questions:   / statisticsglobe  
Facebook – Python Programming Group for Discussions & Questions:   / statisticsglobepython  
Instagram:   / statisticsglobecom  
TikTok:   / statisticsglobe