draw.R

# Drawing a scatter plot of raster images
doInstall <- TRUE  # Change to FALSE if you don't want packages installed.
toInstall <- c("png", "devtools", "MASS", "RCurl")
if(doInstall){install.packages(toInstall, repos = "http://cran.r-project.org")}
lapply(toInstall, library, character.only = TRUE)

# Some helper functions, lineFinder and makeTable
source_gist("818983")
source_gist("818986")

files <- list.files("E:\\Downloads\\yarn")
pngList <- list()

for(filename in files){
  tempPNG <- readJPEG(paste0("E:\\Downloads\\yarn\\",filename))  # Downloads & loads PNGs
  pngList[[filename]] <- tempPNG  # And assigns them to a list.
}

# Very simple dimension reduction -- just the mean R, G, and B values

RBGPos <- t(sapply(pngList, function(ll){
  ll[, , -4][1:3]
}))
RBGPos <- as.data.frame(RBGPos)
meanRGB <- as.data.frame(meanRGB)

PBGPos <- cbind(RBGPos, meanRGB)

colnames(RBGPos) <- c("R", "B", "G")
RBGPos$file <- rownames(RBGPos)

# The dimensions of each item are equal to the pixel dimensions of the .PNG
flagDimensions <- t(sapply(pngList, function(ll){
  dim<-
}))

# Similarity, through Kruskal non-metric MDS
distance <- dist(meanRGB)
distance[distance <= 0] <- 1e-10

MDS <- isoMDS(distance)$points
plot(meanRGB, col = rgb(meanRGB), pch = 20, cex = 2)
meanRGB[,2] <- 1

meanRGB <- t(sapply(pngList, function(ll){
  apply(ll[, , -4], 3, mean)
}))

#RBG to XYZ
for(ii in 1:length(pngList)){ 
  tempName <- rownames(meanRGB)[ii]
  
  #meanRGB[tempName,1] <- gamma.correct(meanRGB[tempName,1])
  #meanRGB[tempName,2] <- gamma.correct(meanRGB[tempName,2])
  #meanRGB[tempName,3] <- gamma.correct(meanRGB[tempName,3])
  
  x <- meanRGB[tempName,1] * 0.649926 + meanRGB[tempName,2] * 0.103455 + meanRGB[tempName,3] * 0.197109
  y <- meanRGB[tempName,1] * 0.234327 + meanRGB[tempName,2] * 0.743075 + meanRGB[tempName,3] * 0.022598
  z <- meanRGB[tempName,1] * 0.0000000 +  meanRGB[tempName,2] * 0.053077 + meanRGB[tempName,3] * 1.035763
  
  meanRGB[tempName, 1] <- x / (x + y + z);
  meanRGB[tempName, 2] <- y / (x + y + z);
}

# Plot:
boxParameter <- 2000  #6000  # To alter dimensions of raster image bounding box
par(bg = gray(8/9))
plot(meanRGB, type = "n", asp = 1)
for(ii in 1:length(pngList)){  # Go through each flag
  tempName <- rownames(meanRGB)[ii]
  Coords <- meanRGB[tempName, 1:2]  # Get coordinates from MDS
  rasterImage(pngList[[tempName]],  # Plot each flag with these boundaries:
              Coords[1]-40/boxParameter, Coords[2]-40/boxParameter,
              Coords[1]+40/boxParameter, Coords[2]+40/boxParameter)
}

boxParameter <- 2000  #6000  # To alter dimensions of raster image bounding box
par(bg = gray(8/9))
plot(meanRGB, type = "n", asp = 1)
for(ii in 1:length(pngList)){  # Go through each flag
  tempName <- rownames(meanRGB)[ii]
  Coords <- meanRGB[tempName, 1:2]  # Get coordinates from MDS
  Dims <- flagDimensions[tempName, ]  # Get pixel dimensions
  rasterImage(pngList[[tempName]],  # Plot each flag with these boundaries:
              Coords[1]-75/boxParameter, Coords[2]-75/boxParameter,
              Coords[1]+75/boxParameter, Coords[2]+75/boxParameter)
}

gamma.correct <- function(color) {
  if(color > 0.04045) {
    return((color + 0.055) / (1.0 + 0.055) ^ 2.4)
  } else {
    return(color / 12.92)
  }
}