1

I am currently working on a problem with R. I'm pretty new to R so I lack of experience, on a (I guess) simple issue. I have a problem with scaling an image in relation to some data I have. The Image is a floor plan. The data I have is recorded manually.

the data looks like this: Data

My Code looks like this:

theData <-data.frame(EntryNr = c(0001,0002,0003,0004,0005,0006,0007), 
                 TimeSet = c('2017-01-15','2017-01-17','2017-01-18','2017-01-19','2017-01-20','2017-01-21','2017-01-22'),
                 SomeID = c('Mario','Mario','Mario','Luigi','Luigi','Luigi','Bowser'),
                 Room = c('Room1','Room1', 'Room1', 'Room1', 'Room1', 'Room1','Room1'),
                 theX = c(12.011, 11.767, 11.715, 11.827, 11.773,11.846,11.781), 
                 theY = c(6.733, 6.698, 6.871, 6.799, 6.887, 6.327,6.577),
                 theZ = c(3, 2.958, 2.983, 2.981, 2.992,2.952,2.945))

thePicture <- readPNG("FloorPlan.png")

ggimage(thePicture, fullpage = FALSE) +
  geom_point(aes(x = theX, y = theY, colour = SomeID), 
             data = theData, size = I(5), fill = NA) +
  labs(x='X axis', y='Y axis')

and my plot looks like this in the end (the background image is a simple floor plan): Plotted image from RStudio

So my problem now is, that the X and Y axis have really high scales. Y goes over 600 and X over 400. But according to the data, the dots should be seen in the plot in "Room1" (bottom right) instead of the bottom left.

Is there a way to rescale the picture?

Something like:

X axis scales from 0 to 15

Y axis scales from 0 to 25

the over-all aim of this, is to plot heatmaps from a lot more data I have, and show which ID was in which "room" for the most time.

EDIT: if I use the ggimage like this:

ggimage(thePicture, fullpage = FALSE, scale_axes = TRUE) +
  geom_point(aes(x = theX, y = theY,  colour = SomeID), 
             data = theData, size = I(5), fill = NA) +
  labs(x='X axis', y='Y axis')

scale_axes = TRUE, my whole image shrinks to the bottom left corner and the scale goes from X = 0 to 12, Y = 0 to 7. so its still not the way i want it.

www
  • 38,575
  • 12
  • 48
  • 84
BlainTheMono
  • 49
  • 9
  • I guess it used the pixels from the png... maybe you have to use a factor to get to the correct positions – drmariod Feb 07 '17 at 15:54
  • how would that look like? like for example? (still pretty new to R). many thanks for the hint. – BlainTheMono Feb 07 '17 at 16:05
  • How about rescaling your `theX` and `theY` to be in the same range as the dimensions of your image? – Artem Sokolov Feb 07 '17 at 16:44
  • Thanks for the advice @ArtemSokolov, but this might work for a few hundred of X and Y but I have like thousands of data sets like this. There has to be a way to put some "grid" on the image or something. – BlainTheMono Feb 07 '17 at 16:46
  • Could You post your systemInfo()... An example fails on my system because of wrong version numbers – drmariod Feb 08 '17 at 05:36

1 Answers1

0

What you need is a scaling factor for each axis! For example you said the x axis ranges from 0 to 15 and the y axis from 0 to 25, the result would look like this...

library(png)
library(ggmap)

thePicture <- readPNG('~/Desktop/Screen Shot 2016-12-01 at 13.03.23.png')
y_factor <- dim(thePicture)[1] / 25
x_factor <- dim(thePicture)[2] / 15

theData <-data.frame(EntryNr = c(0001,0002,0003,0004,0005,0006,0007), 
                     TimeSet = c('2017-01-15','2017-01-17','2017-01-18','2017-01-19','2017-01-20','2017-01-21','2017-01-22'),
                     SomeID = c('Mario','Mario','Mario','Luigi','Luigi','Luigi','Bowser'),
                     Room = c('Room1','Room1', 'Room1', 'Room1', 'Room1', 'Room1','Room1'),
                     theX = c(12.011, 11.767, 11.715, 11.827, 11.773,11.846,11.781), 
                     theY = c(6.733, 6.698, 6.871, 6.799, 6.887, 6.327,6.577),
                     theZ = c(3, 2.958, 2.983, 2.981, 2.992,2.952,2.945))

ggimage(thePicture, fullpage = FALSE) +
  geom_point(aes(x = theX * x_factor , y = theY * y_factor, colour = SomeID), 
             data = theData, size = I(5), fill = NA) +
  labs(x='X axis', y='Y axis')

Just use dim to get the dimensions of your png image. The first number is the y axis, the second the x axis. I don't know what the third number is... Use the ranges you mentioned for each axis and this should do the trick.

drmariod
  • 11,106
  • 16
  • 64
  • 110