CODEWITHSUNDEEP
r
user3777306
user3777306

Reputation: 5

r image(x,y,col=rgb,...) function giving error

I have the following code in R:

img = read.csv(file="img119.csv")

img is loaded like this:

X   Y   R   G   B
1   0   0   0.48046875  0.65234375  0.83593750
2   0   1   0.40234375  0.57031250  0.73828125
3   0   2   0.24609375  0.39453125  0.54687500
4   0   3   0.07031250  0.19921875  0.32031250
5   0   4   0.00000000  0.07812500  0.17187500


width = img[length(img[,1]),1]+1;

height = img[length(img[,2]),2]+1;

x = 1:width;

y = 1:height;

img[,3:5] = img[,3:5]/256;

rgb_colors = rgb(img[,3:5])

imgfunction <-function(xx,yy){

  idx = xx*width + height - yy;

  rgb_color = rgb_colors[idx];

  return ( rgb_color )

} 

rgb_colors_xy = outer(x,y,imgfunction)

image(x,y,rgb_colors)

I want to draw that csv file like a regular image. However, the last line gives me error as following: 

Error in image.default(x, y, rgb_colors) : 'z' must be a matrix

I am not sure what I am doing wrong and appreciate if you help me fix this code.

Thanks,

Upvotes: 0

Views: 1605

Answers (1)

Carl Witthoft
Carl Witthoft

Reputation: 21502

graphics::image will not create a color image from 3-D data. You need to use raster::image or rasterImage . Read the help files carefully to understand how your RGB data must be converted to arrays. The "long" way is to create three x by y matrices and fill each one with something along the lines of

for(i in 1:max(x)) {
    for(j in 1:max(y)) redmat[i,j]<- img[i,j,3]
}

There are faster tools which someone more awake than I will probably post. Note also that R indexes from 1, not zero, so you'll need to offset the counts appropriately.

Upvotes: 1

Related Questions