artist-frequency-raster

artist-frequency-raster.R

require(stats)
require(ggplot2)

# Download your last.fm history as a .csv from here: 
# https://benjaminbenben.com/lastfm-to-csv/

# load 'em
lfmdat <- read.csv('~/lastfm.csv', header = FALSE)
names(lfmdat) <- c("artist", "album", "song", "time")

# clean blanks like a damn animal
lfmdat <- lfmdat[lfmdat[,]$artist != "",]
lfmdat <- lfmdat[lfmdat[,]$album != "",]
lfmdat <- lfmdat[lfmdat[,]$song != "",]
lfmdat <- lfmdat[lfmdat[,]$time != "",]

# Format the timestamps, see ?strptime
lfmdat$time <- strptime(lfmdat$time, format = "%d %b %Y %R")

# Assume the artists/albums are well formed and factorize
lfmdat$artist <- as.factor(lfmdat$artist)
lfmdat$album  <- as.factor(lfmdat$album)

#####
# Frequency raster plot

# order artist factor by frequency
ftab = table(lfmdat$artist)

# pretty janky and slow but stfu
lfmdat$freq <- 0
for (a in names(ftab)){
  lfmdat[lfmdat$artist == a,]$freq <- ftab[a]
}

lfmdat$artist <- factor(lfmdat$artist, levels=lfmdat$artist[order(lfmdat$freq)], ordered=TRUE)

# now plot that shit
g <- ggplot(lfmdat, aes(x=time, y=artist))+
  geom_point(size=0.05)+
  scale_x_datetime(date_breaks = "1 month", date_labels = "%m-%y")+
  theme(
    axis.text.y = element_blank(),
    axis.title.y = element_blank(),
    axis.ticks.y = element_blank(),
    axis.title.x = element_blank(),
    axis.ticks.x = element_blank(),
    panel.grid.major = element_blank(),
    panel.grid.minor = element_blank(),
    panel.background = element_blank(),
    legend.position = "none",
    axis.text.x = element_text(angle=90, hjust = 1))

dir.create("~/wastin_ur_harddrive")
ggsave(paste("~/wastin_ur_harddrive/on_this_stupid_shit_at_",as.numeric(as.POSIXct(Sys.time())),".png",sep=""),
  plot=g, device="png", width=4, height=7, units="in", dpi=500, bg="transparent")