[solved] Mouse click to return grid coordinates instead

Comments

• Edwin

  September 2012 edited September 2012

  Hi Kyle!

  If I understand correctly, you want to register a mouse click position with regards to your picture. Is that correct?

  In that case, you might want to remove the mouse_response and replace it with an inline_script. In the run phase, you can type the following:

  from openexp.canvas import canvas
  from openexp.mouse import mouse
  
  # FILL THIS OUT YOURSELF
  picsize = (400,200) # width and height in pixels
  sketchpadname = "my_sketchpad" # name of the sketchpad in which you present your picture
  
  # create items
  my_canvas = canvas(self.experiment)
  my_mouse = mouse(self.experiment)
  
  # copy your sketchpad to canvas
  my_canvas.copy(exp.items[sketchpadname].canvas)
  
  # show canvas
  my_canvas.show()
  t0 = self.experiment.time()
  
  # wait for mouse response
  no_click = True
  while no_click:
      # get mouse position
      pos, postime = my_mouse.get_pos()
      # draw cursor
      my_canvas.fixdot(x=pos[0],y=pos[1])
      my_canvas.show()
      # wait for buttonpress
      click, clickpos, clicktime = my_mouse.get_click(timeout=5)
      # if button is clicked...
      if click:
          RT = clicktime - t0 # ...record time and...
          # ...check if the buttonclick is in between the picture's left and right edges and...
          if clickpos[0] > (self.get("width")/2) - (picsize[0]/2) and clickpos[0] < (self.get("width")/2) + (picsize[0]/2):
              # ...check if the buttonclick is in between the picture's top and bottom edges
              if clickpos[1] > (self.get("height")/2) - (picsize[1]/2) and clickpos[1] < (self.get("height")/2) + (picsize[1]/2):
                  no_click = False # discontinue while loop
      # reset screen (remove cursor)
      my_canvas.copy(exp.items[sketchpadname].canvas)
      
  # recalculate position
  response_x = clickpos[0] + self.get("width")/2
  response_y = clickpos[1] + self.get("height")/2
  
  # save some stuff
  self.experiment.set("response_x", response_x)
  self.experiment.set("response_y", response_y)
  self.experiment.set("response_time", RT)
  

  Since a standard OpenSesame canvas uses the top left of the screen as it's origin (point (0,0)), the only thing you have to do is add the distance to where you want your origin to be. In the above script, it is the center of the screen (and thus the center of your picture). I do not know if this is indeed the location that you wanted to calculate, so see below for the other options (replace the code underneath '# recalculate position' with one of the following):

  (0,0) is top left of picture:

  # recalculate position
  response_x = clickpos[0] + (self.get("width")/2 - (picsize[0]/2))
  response_y = clickpos[1] + (self.get("height")/2 - (picsize[1]/2))
  

  (0,0) is bottom left of picture:

  # recalculate position
  response_x = clickpos[0] + (self.get("width")/2 - (picsize[0]/2))
  response_y = clickpos[1] + (self.get("height")/2 + (picsize[1]/2))
  

  Please note that you should set your sketchpad's duration to 0 and that you should adjust the values of picsize and sketchpadname yourself!

  Good luck!
  Edwin
• Kyle

  September 2012 edited 10:08PM

  Edwin,

  Thank you for the inline-script direction and code. This seems to function well, but I'm still not getting the correct coordinates. I have the picture centered in the sketchpad (which I wonder, does the grid in the sketchpad have any bearing on the pictures location or is it simply a visual reference?) Still, if I run my experiment and click on the center of the picture I return my maximum resolution tuple 1280,800 (x,y).

  Looking at your code, I see that you are dividing the length and width in half to reach the center of the picture, so that seems clear. I have added the correct picture size at the top of course (774,689).

  Any ideas? This is so close, I feel it may be a bit of code work and we've got it. I'll tinker more with it as well.

  Thanks again,

  Kyle
• Edwin

  September 2012 edited September 2012

  Woops, my bad! To recalculate the coordinates to the cenre one shouldn't add half of the screen size, but deduct it in stead. I had just disected a human brain, so I was a bit distracted... Sorry!
• Kyle

  September 2012 edited 10:08PM

  Perfect! That was the chink in the armour. No we're rock solid. Thank you again sir!
• tuongvan

  October 2014 edited October 2014

  Hi,
  I would like to ask then how to devide areas on a picture into different regions and then register the mouse clicks of the participants in these pre-defines regions? So I basically need the location where participants click on the picture, not in absolute coordinates but in terms of specific regions. Thank you very much!
• sebastiaan

  October 2014 edited 10:08PM

  I would like to ask then how to devide areas on a picture into different regions and then register the mouse clicks of the participants in these pre-defines regions?

  This depends on how you have defined your regions of interest (ROIs). If you simple divide the screen into a grid, you can use the touch_response plug-in, as described here:

  • http://osdoc.cogsci.nl/tutorials/capybara/#step-10-define-the-touch-response

  If you have defined ROIs in a different way, then you first need to know exactly how they are defined. For example, do you have a spreadsheet with coordinates and radii that define your ROIs? Or do you have a bitmap in which ROIs are color coded? Or something different altogether?

  
  
  Check out SigmundAI.eu for our OpenSesame AI assistant!