def getScreenMeasurements(dpath, et_model, et_model_display_configs):
"""
:param dpath:
:param et_model:
:param et_model_display_configs:
:return:
"""
# et_model, _ = getInfoFromPath(dpath)
display_param = et_model_display_configs.get(et_model)
if display_param is None:
et_config_path = glob.glob('./configs/*%s.yaml' % (et_model))
if et_config_path:
et_config_path = nabs(et_config_path[0])
display_config = load(file(et_config_path, 'r'), Loader=Loader)
dev_list = display_config.get('monitor_devices')
for d in dev_list:
if d.keys()[0] == 'Display':
d = d['Display']
width = d.get('physical_dimensions', {}).get('width')
height = d.get('physical_dimensions', {}).get('height')
eye_dist = d.get('default_eye_distance', {}).get(
'surface_center')
et_model_display_configs[et_model] = OrderedDict()
et_model_display_configs[et_model][
screen_measure_fields[0]] = width
et_model_display_configs[et_model][
screen_measure_fields[1]] = height
et_model_display_configs[et_model][
screen_measure_fields[2]] = eye_dist
return et_model_display_configs[et_model]
return display_param
def getGeometry(data):
return np.mean((1/(np.degrees(2*np.arctan(data['screen_width']/(2*data['eye_distance'])))/data['display_width_pix']),
1/(np.degrees(2*np.arctan(data['screen_height']/(2*data['eye_distance'])))/data['display_height_pix'])))
if not os.path.exists(OUTPUT_DIR):
os.mkdir(OUTPUT_DIR)
OUTPUT_DIR = getFullOutputFolderPath(OUTPUT_DIR)
if not os.path.exists(OUTPUT_DIR):
os.mkdir(OUTPUT_DIR)
print 'OUTPUT_FOLDER:', OUTPUT_DIR
DATA_FILES = [nabs(fpath) for fpath in glob.glob(GLOB_PATH_PATTERN) if analyseit(fpath, INCLUDE_SUB, INCLUDE_TRACKERS)]
stim_all = []
if __name__ == '__main__':
for file_path in DATA_FILES:
file_log = open(OUTPUT_DIR + '/win_select.log', 'a')
try:
# if 1:
t1 = time.time()
DATA = np.load(file_path)
et_model, sub = getInfoFromPath(file_path)
#Removes every second sample for LC Tech EyeFollower
#TODO: come up with a better alternative
'min_cal_points': 8,
'win_select_func': 'roll',
'win_size': 0.175,
'win_type': 'sample',
'window_skip': 0.2,
'wsa': 'fiona' ,
'units': 'gaze',
}]
if not os.path.exists(OUTPUT_FOLDER):
os.mkdir(OUTPUT_FOLDER)
OUTPUT_FOLDER = getFullOutputFolderPath(OUTPUT_FOLDER)
print 'OUTPUT_FOLDER:', OUTPUT_FOLDER
DATA_FILES = [nabs(fpath) for fpath in glob.glob(GLOB_PATH_PATTERN) if
analyseit(fpath)]
#Check for dublicates
sub_dict = dict()
for et_model in INCLUDE_TRACKERS:
sub_dict[et_model] = []
for file_path in DATA_FILES:
et, sub = getInfoFromPath(file_path)
if et==et_model:
sub_dict[et_model].append(sub)
if len(sub_dict[et_model]) != len(np.unique(sub_dict[et_model])):
unique_ids, unique_counts = np.unique(sub_dict[et_model], return_counts=True)
print 'Dublicate subject ids in %s:'%et_model, \
unique_ids[np.argwhere(unique_counts > 1).flatten()]