MATLAB: How to initiate a while loop with a GUI button (using UI control) and then end that loop with a different GUI button

matlab guiserialuicontrolwhile loop

I am working with a tracking system and I have been using a while loop to continuously scan for data. The whole time I have been terminating the loop with control-c, but now I am manipulating an old GUI and trying to initiate and terminate the loop with buttons in the GUI. I thought this might work but I am not having any luck:
%pivot shift button %collects data while subject performs pivot shift PS_data = uicontrol('Position',[1125 250 100 25],'String','Pivot Shift', 'FontSize',12,'Callback','GUI_data_collection'); %calls data collection
%This ends the pivot shift PivotShiftEndButton = uicontrol('Position',[1125 200 100 25],'String','End Shift',…, 'FontSize',12,'Callback','delete(PS_data)');
fprintf(Ftrak,'C'); shift_data = [];
while ishandle(PS_data)
scan = fscanf(Ftrak,'%f');
scan= scan';
if length(scan) == 7
shift_data = [shift_data;scan];
if shift_data(end,1) == shift_data(end - 1,1)
shift_data(end,:) = [];
end
end
end

Best Answer

Do not use a string to define a callback. This is supported for backward compatibility with the ancient Matlab 5.3 only. Use a function handle instead:
...'Callback', @GUI_data_collection
'delete(PS_data)' is a bad callback also: Remember that callbacks, which are defined as strings, are evaluated in the base workspace. There PS_data is unknown. Better use a flag stored in the GUI:
PivotShiftEndButton = uicontrol('Position',[1125 200 100 25],'String','End Shift',...,
'Callback', @StopDataCollection, ...
'UserData' 1); % <-- The flag
fprintf(Ftrak,'C');
shift_data = [];
while get(PivotShiftEndButton, 'UserData')
scan = fscanf(Ftrak,'%f').';
if length(scan) == 7
shift_data = [shift_data;scan];
if shift_data(end,1) == shift_data(end - 1,1)
shift_data(end,:) = [];
end
end
drawnow; % Give the GUI a chance to update
end
function StopDataCollection(hObject, EventData)
set(hObject, 'UserData', 0);
end
Note that the iterative growing of the array uses a huge amount of resources. In each iteration 7 elements a 8 bytes are appended. Therefore a new array is created an the old values are copied. for 10'000 iterations, this does not allocate 10'000*56 bytes=560kB, but sum(1:10000) * 56 bytes: 2.8 GB!
Better pre-allocate the output with a maximum number of recorded data. Although such a limit is a drawback, the current method is limited by exhausting the computer until it crashs.