Use ptb's windowsize instead of matlab's in screen fields initialization.

Typo fix in the fixation generator doc.

Author: Genuster

+CFSVM/+Element/+Screen/CustomScreen.m

@@ -44,13 +44,13 @@
             obj.background_color = kwargs.background_color;
             obj.fields{1} = CFSVM.Element.Screen.ScreenField(kwargs.initial_rect);
             if kwargs.is_stereo
-                set(0, 'units', 'pixels');
-                screen_size = get(0, 'ScreenSize');
+                [screen_width, ~] = Screen('WindowSize', max(Screen('Screens')));
                 second_rect = [ ...
-                               screen_size(3) - kwargs.initial_rect(3), ...
+                               screen_width - kwargs.initial_rect(3), ...
                                kwargs.initial_rect(2), ...
-                               screen_size(3) - kwargs.initial_rect(1), ...
+                               screen_width - kwargs.initial_rect(1), ...
                                kwargs.initial_rect(4)];
+                
                 obj.fields{2} = CFSVM.Element.Screen.ScreenField(second_rect);
             end
             obj.shift = kwargs.shift;

docs/fixation_generator.md

@@ -13,7 +13,7 @@ import CFSVM.Generators.FixationGenerator
 ```
 
 ### Initialize the generator object
-The arguments of the object constructor are as follows: first, the path to the parent directory in which the `Fixation/` folder will be created. The generated fixation target will be saved inside this folder. Next, the **radius** argument describes the radius in pixels of the largest shape in the fixation, e.g., radius of the big circle in the "AB" shape or half-width of the cross in "AC" shape, the final image resolution will be (radius\*2+1, radius\*2+1).We suggest using a minimum of 128 pixels and resizing it later in the experiment code, instead of generating an image with low resolution.
+The arguments of the object constructor are as follows: first, the path to the parent directory in which the `Fixation/` folder will be created. The generated fixation target will be saved inside this folder. Next, the **radius** argument describes the radius in pixels of the largest shape in the fixation, e.g., radius of the big circle in the "AB" shape or half-width of the cross in "AC" shape, the final image resolution will be (radius\*2+1, radius\*2+1). We suggest using a minimum of 128 pixels and resizing it later in the experiment code, instead of generating an image with low resolution.
 
 After that, the **hex_color** sets the color of the fixation target. If you'd like to smooth the edges of the image, set **is_smooth_edges** to `true` and **smoothing_cycles** to some positive integer (the greater the number, the smoother edges will be). You can also set **is_outline** to `true` to leave only outlines of the fixation (you will need an [installed](https://uk.mathworks.com/matlabcentral/answers/101885-how-do-i-install-additional-toolboxes-into-an-existing-installation-of-matlab#answer_111232) Image Processing Toolbox for outlining).