Last updated on 2026-04-04
---
title: "Artists Analysis"
description: "Top artists by month and total listening statistics"
---
```{r setup}
#| message: false
#| warning: false
library(here)
source(here("src", "utils", "utils.R"))
load_common_libraries()
library(viridis)
```
```{r load-artist-data}
#| message: false
#| warning: false
all_artist_data <- load_artist_data()
```
## Top 5 Artists — This Month, Last Month, and Historical
```{r top-artists-combined}
#| message: false
current_month <- floor_date(Sys.Date(), "month")
last_month <- current_month - months(1)
top5 <- function(data, n = 5) {
data %>%
select(name, date) %>%
distinct() %>%
count(name, sort = TRUE) %>%
head(n) %>%
mutate(rank = row_number()) %>%
select(rank, artist = name, plays = n) %>%
mutate(combined = paste0(artist, "\n(", plays, " days)"))
}
top_5_current_month <- all_artist_data %>% filter(year_month == current_month) %>% top5()
top_5_last_month <- all_artist_data %>% filter(year_month == last_month) %>% top5()
top_5_historical <- all_artist_data %>% top5()
combined_table <- data.frame(
Rank = 1:5,
`This Month` = top_5_current_month$combined[match(1:5, top_5_current_month$rank)],
`Last Month` = top_5_last_month$combined[match(1:5, top_5_last_month$rank)],
`Historical` = top_5_historical$combined[match(1:5, top_5_historical$rank)],
check.names = FALSE
)
combined_table[is.na(combined_table)] <- "-"
combined_table %>%
gt() %>%
tab_header(
title = "Top 5 Artists Comparison",
subtitle = paste("This Month:", format(current_month, "%B %Y"),
"| Last Month:", format(last_month, "%B %Y"),
"| Historical: All Time")
) %>%
cols_label(
Rank = "Rank",
`This Month` = paste("This Month", format(current_month, "(%B %Y)")),
`Last Month` = paste("Last Month", format(last_month, "(%B %Y)")),
`Historical` = "Historical (All Time)"
) %>%
opt_interactive(use_search = FALSE)
```
## Top 5 Artists by Total Songs and Days Listened
```{r top-artists-pivoted}
#| message: false
top_5_total_songs <- all_artist_data %>%
count(name, sort = TRUE) %>%
head(5) %>%
mutate(rank = row_number()) %>%
select(rank, artist = name, total_songs = n)
top_5_total_days <- all_artist_data %>%
select(name, date) %>%
distinct() %>%
count(name, sort = TRUE) %>%
head(5) %>%
mutate(rank = row_number()) %>%
select(rank, artist = name, total_days = n)
get_artist_info <- function(artist_df, rank_num) {
if (nrow(artist_df) >= rank_num && !is.na(artist_df$artist[rank_num])) {
if ("total_songs" %in% names(artist_df)) {
paste0(artist_df$artist[rank_num], " (", artist_df$total_songs[rank_num], ")")
} else {
paste0(artist_df$artist[rank_num], " (", artist_df$total_days[rank_num], ")")
}
} else {
"-"
}
}
pivoted_table <- data.frame(
Metric = c("Total Songs", "Total Days"),
`Rank 1` = c(get_artist_info(top_5_total_songs, 1), get_artist_info(top_5_total_days, 1)),
`Rank 2` = c(get_artist_info(top_5_total_songs, 2), get_artist_info(top_5_total_days, 2)),
`Rank 3` = c(get_artist_info(top_5_total_songs, 3), get_artist_info(top_5_total_days, 3)),
`Rank 4` = c(get_artist_info(top_5_total_songs, 4), get_artist_info(top_5_total_days, 4)),
`Rank 5` = c(get_artist_info(top_5_total_songs, 5), get_artist_info(top_5_total_days, 5)),
check.names = FALSE
)
pivoted_table %>%
gt() %>%
tab_header(
title = "Top 5 Artists by Total Songs and Days Listened",
subtitle = "Metrics as rows, top 5 artists as columns (Artist Name (KPI))"
) %>%
cols_label(Metric = "Metric") %>%
opt_interactive(use_search = FALSE)
```
## Top Artists Trend Over Time
```{r top-artists-trend}
#| fig-width: 14
#| fig-height: 6
# Monthly play counts for the top 10 all-time artists
top10_artists <- all_artist_data %>%
count(name, sort = TRUE) %>%
head(10) %>%
pull(name)
artist_monthly <- all_artist_data %>%
filter(name %in% top10_artists) %>%
count(name, year_month, name = "plays") %>%
arrange(year_month)
p <- artist_monthly %>%
ggplot(aes(x = year_month, y = plays, color = name, group = name)) +
geom_line(linewidth = 1, alpha = 0.7) +
geom_point_interactive(
aes(
tooltip = paste0("<b>", name, "</b><br>", format(year_month, "%B %Y"), "<br>Plays: ", plays),
data_id = paste(name, year_month, sep = "-")
),
size = 2
) +
scale_color_viridis_d(option = "turbo", name = "Artist") +
scale_x_date(date_labels = "%b %Y", date_breaks = "6 months") +
labs(
title = "Monthly Plays — Top 10 Artists",
subtitle = "How often you listened to each top artist over time",
x = NULL,
y = "Songs Played"
) +
theme_minimal() +
theme(
plot.title = element_text(size = 16, face = "bold"),
plot.subtitle = element_text(size = 12, color = "gray60"),
axis.text.x = element_text(angle = 45, hjust = 1),
legend.position = "right"
)
make_girafe(p, height_svg = 6)
```
## Artist Diversity Over Time
```{r artist-diversity}
#| fig-width: 14
#| fig-height: 5
diversity_data <- all_artist_data %>%
count(date, name = "n_artists_day") %>%
arrange(date) %>%
mutate(rolling_avg = zoo::rollmean(n_artists_day, k = 30, fill = NA, align = "right"))
if (requireNamespace("zoo", quietly = TRUE)) {
p <- diversity_data %>%
ggplot(aes(x = date)) +
geom_line(aes(y = n_artists_day), color = "#4A90E2", alpha = 0.3, linewidth = 0.5) +
geom_line_interactive(
aes(
y = rolling_avg,
tooltip = paste0(format(date, "%b %d, %Y"), "<br>30-day avg artists/day: ", round(rolling_avg, 1)),
data_id = as.character(date)
),
color = "#1DB954", linewidth = 1.5, na.rm = TRUE
) +
labs(
title = "Artist Diversity Over Time",
subtitle = "Unique artists per day (blue) and 30-day rolling average (green)",
x = NULL,
y = "Unique Artists per Day"
) +
scale_x_date(date_labels = "%b %Y", date_breaks = "6 months") +
theme_minimal() +
theme(
plot.title = element_text(size = 16, face = "bold"),
plot.subtitle = element_text(size = 12, color = "gray60"),
axis.text.x = element_text(angle = 45, hjust = 1)
)
make_girafe(p, height_svg = 5)
} else {
cat("Install the 'zoo' package to see the rolling average chart.")
}
```
---
*Last updated on `r Sys.Date()`*