#!/usr/bin/env python

#The most accurate way to time your stimulus presentation is to
#present for a certain number of frames. For that to work you need
#your window flips to synchronise to the monitor and not to drop
#any frames. This script examines the precision of your frame flips.

#Shut down as many applications as possible, especially those that
#might try to update

from psychopy import visual, logging, core, event
#visual.useFBO=True

import matplotlib
#matplotlib.use('WXAgg')#change this to control the plotting 'back end'
import pylab

nIntervals=700
win = visual.Window([1280,1024], fullscr=True, allowGUI=False, waitBlanking=True)
progBar = visual.PatchStim(win, tex=None, mask=None,
    size=[0,0.05],color='red',pos=[0,-0.9],
    autoLog=False)
myStim = visual.PatchStim(win, tex='sin', mask='gauss',
    size=300,sf=0.05, units='pix',
    autoLog=False)
#logging.console.setLevel(logging.INFO)#uncomment if you want to print every frame time


for frameN in range(nIntervals+1):
    if frameN==200:
        win.setRecordFrameIntervals()
    if frameN>200:
        progBar.setSize([2.0*frameN/nIntervals, 0.05])
        progBar.draw()
        myStim.setPhase(0.1,'+')
        myStim.draw()
        if event.getKeys():
            print 'stopped early'
            break
        win.logOnFlip(msg='frame=%i' %frameN, level=logging.EXP)
    win.flip()
win.close()

#calculate some values
intervalsMS = pylab.array(win.frameIntervals)*1000
m=pylab.mean(intervalsMS)
sd=pylab.std(intervalsMS)
# se=sd/pylab.sqrt(len(intervalsMS)) # for CI of the mean
distString= "Mean=%.1fms, s.d.=%.2f, 99%%CI(frame)=%.2f-%.2f" %(m,sd,m-2.58*sd,m+2.58*sd)
nTotal=len(intervalsMS)
nDropped=sum(intervalsMS>(1.5*m))
droppedString = "Dropped/Frames = %i/%i = %.3f%%" %(nDropped,nTotal, 100*nDropped/float(nTotal))

#plot the frameintervals
pylab.figure(figsize=[12,8])
pylab.subplot(1,2,1)
pylab.plot(intervalsMS, '-')
pylab.ylabel('t (ms)')
pylab.xlabel('frame N')
pylab.title(droppedString)
#
pylab.subplot(1,2,2)
pylab.hist(intervalsMS, 50, normed=0, histtype='stepfilled')
pylab.xlabel('t (ms)')
pylab.ylabel('n frames')
pylab.title(distString)
pylab.show()