photoColorAnalysis.R

# Personal Photo Histry
library("rPlotter")
library("magrittr")
library("grid")
library("gridSVG")
library("ggplot2")

#retreives list of every .jpeg on my hard-drive
files_list1 <- list.files(path="//Volumes//Seagate Backup Plus Drive/Pictures/", pattern = ".*(jpg|JPG|jpeg|JPEG)", full.names=T, recursive=T)
files_list2 <- list.files(path="//Volumes//Seagate Backup Plus Drive/Picuters /", pattern = ".*(jpg|JPG|jpeg|JPEG)", full.names=T, recursive=T)

all_jpgs <- c(files_list1, files_list2)

#extracts the file path and creation date
info <- file.info(all_jpgs)
info_path <- info
info_path["file_path"] <- rownames(info)

#extracts the most common color from each photo. This took a while...
pic_colors <- data.frame()
for (i in 1:length(info_path$file_path)) {
  pic_color <- color(info_path$file_path[i])
  pic_colors <- rbind(pic_colors, t(data.frame(pic_color)))
  print(i)
}

#combies the files information with the most common color
per_photo_color <- cbind(info_path, pic_colors)
write.csv(per_photo_color, file = "brett_photo_color.csv")

#----- read data in ---
per_photo_color <- read.csv("Color_Personal/brett_photo_color.csv")

b_photo_color <- per_photo_color[c("mtime", "V1")]
colnames(b_photo_color) <- c("create_date", "color")
b_photo_color <- b_photo_color[order(b_photo_color$create_date),]

#reformats the year coloum
year <- strftime(b_photo_color$create_dat, format="%Y") %>% data.frame()

b_photo_color <- cbind(b_photo_color, year)
b_photo_color["color"] <- as.character(b_photo_color$color)
b_photo_color["."] <- as.character(b_photo_color$.)
colnames(b_photo_color) <- c("create_date", "color", "year")

p_color <- subset(b_photo_color, b_photo_color$color != "NULL")

#plotting the distribution of color

#distribtion of color by year
grid.newpage()
xpos <- .05
for (p in 2006:2015) {
  grid.raster(unlist(subset(p_color, year == p, select = "color")),
              x = xpos,
              width = .1,
              height = .8, 
              interpolate = FALSE)
  grid.text(p, x = xpos, y = .075, gp = gpar(col = "black", fontsize = 13))
  xpos <- xpos + .1
}
grid.text("Personal Photo Collection", x =.5, y = .95, just = c("centre","centre"), gp = gpar(col = "black", fontsize = 30)) 
#exporting data visualization
grid.export("Color_Personal/colorYear.svg")

# number of photos by year
yearFreq <- table(p_color$year) %>% data.frame()
colnames(yearFreq) <- c("year", "freq")
yearFreq$year <- factor(yearFreq$year, levels = yearFreq$year[order(-as.numeric(yearFreq$year))])

yearFreqPlot <- ggplot(yearFreq, aes(x = year, y = freq)) + 
                  geom_bar(stat = "identity", fill = "royal blue") +
                  coord_flip() +
                  labs(title = "Number of photos taken by year", y = "# of Photos", x = "Year") +
                  theme(legend.position="bottom", legend.direction="horizontal") + 
                  scale_y_continuous(limits = c(0,4000), breaks = pretty(0:4000, n = 10), labels = pretty(0:4000, n = 10)) +
                  themeBrettrics

#crearting the svg image
svg(filename = "Color_Personal/yearFreq.svg", 
    width = 7, 
    height = 5, 
    pointsize = 5)
print(yearFreqPlot)
dev.off()

#---functions---

color <- function(x) {
  return(tryCatch({
    extract_colours(x, 1)
  },
  error = function(e) {
    "NULL"
  }))
}