def combine_strokes_if_low_density(self, density_limit=10):
all_x, all_y = [l.get_x() for l in self.strokes], [l.get_y() for l in self.strokes]
all_x, all_y = sum(all_x, []), sum(all_y, [])
full_letter = Stroke(all_x, all_y)
full_pen_up = [compute_pen_up(letter.get_x(), letter.get_y(), density_limit = 10) for letter in self.strokes]
full_pen_up = sum(full_pen_up, [])
new_strokes_idx = self.strokes_idx(full_pen_up)
x_pts = [full_letter.get_x()[idx_start:idx_end] for (idx_start, idx_end)
in zip(new_strokes_idx[:-1], new_strokes_idx[1:])]
y_pts = [full_letter.get_y()[idx_start:idx_end] for (idx_start, idx_end)
in zip(new_strokes_idx[:-1], new_strokes_idx[1:])]
self.strokes = [Stroke(x,y) for x,y in zip(x_pts, y_pts)]
self.uniformize_strokes_with_step(step = 0.1)
self.compute_letter_properties()
def draw_points(self, layer_number, style, points):
# Corrective constants to match characteristics of OpenCanvas 1.1 brushes
pressure_correction_weight = 1.0
brush_size_weight = 0.7
brush_velocity = 0.03
print "Drawing on layer ", layer_number
self.select_layer(layer_number)
layer = self.get_current_layer()
brush_style = BrushInfo(brushdata)
# Set color
rgb = map(lambda c: float(c) / 255.0, style['color'])
brush_style.set_color_rgb(rgb)
# Set size adjustments and calculate pressure curve
size = style['size']
pressure_correct = style['pressure_correct']
x_points = (0.000000, 0.237952, 0.500000, 0.765060, 1.000000)
y_points = map(
lambda x: (x**pressure_correct) * pressure_correction_weight -
pressure_correction_weight / 2.0, x_points)
pressure_curve = zip(x_points, y_points)
brush_style.set_setting(
'radius_logarithmic',
[math.log(size * brush_size_weight), {
'pressure': pressure_curve
}])
brush = Brush(brush_style)
stroke = Stroke()
stroke.start_recording(brush)
tdelta = 0
lx, ly, lp = points[0]
for point in points:
lx = (lx + point[0]) / 2
ly = (ly + point[1]) / 2
lp = float(point[2]) / 1024.0 # Pressure
stroke.record_event(tdelta, lx, ly, lp, 0, 0)
tdelta += brush_velocity
stroke.stop_recording()
# Render stroke and save undo information
snapshot_before = layer.save_snapshot()
stroke.render(layer._surface)
self.do(command.Stroke(self, stroke, snapshot_before))
def __init__(self, x, y):
"""
Provided only x an y, they are expected to be a list of list of coordinates
e.g. x = [[0,1,2], [3,4,5], [6,7,8]], y = [[9,10,11], [12,13,14], [15,16,17]],
pen_up = None
Provided x, y and pen_up, x and y are expected to be a list of coordinates
and pen_up a list of 0 and 1 where 1 means the pen was removed from the
tablet at the considered index
e.g. x = [0,1,2,3,4,5,6,7,8], y = [9,10,11,12,13,14,15,16,17]
"""
self.x_list, self.y_list = x, y
self.x_list = [x for x in self.x_list if len(x)>0]
self.y_list = [x for x in self.y_list if len(x)>0]
self.strokes = [Stroke(x_sublist, y_sublist) for x_sublist, y_sublist in zip(self.x_list, self.y_list)
if len(x_sublist)>0 and len(y_sublist)>0]
self.original_xmin = min([min(sublist) for sublist in self.x_list]) if(len(self.strokes)>0) else 0
self.original_xmax = max([max(sublist) for sublist in self.x_list]) if(len(self.strokes)>0) else 0
self.original_ymin = min([min(sublist) for sublist in self.y_list]) if(len(self.strokes)>0) else 0
self.original_ymax = max([max(sublist) for sublist in self.y_list]) if(len(self.strokes)>0) else 0
self.original_width = (self.original_xmax - self.original_xmin) if(len(self.strokes)>0) else 0
self.original_height = (self.original_ymax - self.original_ymin) if(len(self.strokes)>0) else 0
self.xmin, self.xmax, self.ymin, self.ymax, self.width, self.height = 0, 0, 0, 0, 0, 0
self.compute_letter_properties()
def stroke(self, idx_stroke):
if idx_stroke<len(self.strokes):
return self.strokes[idx_stroke]
else:
print("Invalid stroke id ({}), max id is {}".format(idx_stroke, len(self.strokes)))
print("Returning empty stroke")
return Stroke()
def to_stroke_with_pen_up(self):
x = sum([stroke.get_x() for stroke in self.strokes], [])
y = sum([stroke.get_y() for stroke in self.strokes], [])
pen_ups = [[0 for i in range(len(stroke.get_x()))] for stroke in self.strokes]
for pen_up in pen_ups:
pen_up[0] = 1
pen_ups = sum(pen_ups, [])
return Stroke(x,y), pen_ups
def uniformize_strokes_with_step(self, step = 0.1):
strokes = list()
for stroke in self.strokes:
if(len(stroke.get_x())<=1 and len(stroke.get_y())<=1):
continue
x_pts, y_pts = stroke.get_x(), stroke.get_y()
densified_stroke = [evenly_spaced_interpolation(x1,y1,x2,y2) for x1, y1, x2, y2
in zip(x_pts[:-1], y_pts[:-1], x_pts[1:], y_pts[1:])]
x, y = [s["x"] for s in densified_stroke], [s["y"] for s in densified_stroke]
x, y = sum(x, []), sum(y, [])
strokes.append(Stroke(x,y))
self.strokes = strokes
def draw_points(self, layer_number, style, points):
# Corrective constants to match characteristics of OpenCanvas 1.1 brushes
pressure_correction_weight = 1.0
brush_size_weight = 0.7
brush_velocity = 0.03
print "Drawing on layer ", layer_number
self.select_layer(layer_number)
layer = self.get_current_layer()
brush_style = BrushInfo(brushdata)
# Set color
rgb = map(lambda c: float(c) / 255.0, style['color'])
brush_style.set_color_rgb(rgb)
# Set size adjustments and calculate pressure curve
size = style['size']
pressure_correct = style['pressure_correct']
x_points = (0.000000, 0.237952, 0.500000, 0.765060, 1.000000)
y_points = map(lambda x: (x**pressure_correct) * pressure_correction_weight - pressure_correction_weight / 2.0, x_points)
pressure_curve = zip(x_points, y_points)
brush_style.set_setting('radius_logarithmic', [math.log(size*brush_size_weight), {'pressure': pressure_curve}])
brush = Brush(brush_style)
stroke = Stroke()
stroke.start_recording(brush)
tdelta = 0
lx, ly, lp = points[0]
for point in points:
lx = (lx + point[0]) / 2
ly = (ly + point[1]) / 2
lp = float(point[2]) / 1024.0 # Pressure
stroke.record_event(tdelta, lx, ly, lp, 0, 0)
tdelta += brush_velocity
stroke.stop_recording()
# Render stroke and save undo information
snapshot_before = layer.save_snapshot()
stroke.render(layer._surface)
self.do(command.Stroke(self, stroke, snapshot_before))
def loading_traces(self, root, debug=False):
"""
extract all traces from Inkml File
generate scaled coordinates with H is maxed up to 10000
"""
traces = dict()
for child in root:
tag = child.tag
attrib = child.attrib
#print child.tag, child.attrib
if tag.endswith(
'annotation') and attrib and attrib['type'] == 'truth':
self.latex = child.text.strip()
if child.tag[-5:] == 'trace':
trace_id = child.attrib['id']
coords = child.text
splitted = coords.split(',')
coords = np.zeros((2, len(splitted)))
idx = 0
for x_y in splitted:
x_y = x_y.strip().split(' ')
coords[:, idx] = [float(x_y[0]), float(x_y[1])]
idx += 1
trace_id_int = int(trace_id)
self.id_count.next_id = trace_id_int
traces[trace_id_int] = Stroke(trace_id_int, coords)
if debug:
print('trace dict:\n', traces)
# //Compute bounding box of the input expression
x_min, y_min, x_max, y_max = get_bounding_box(
traces) # bounding box for the whole math expression
# Just in case there is only one point or a sequence of points perfectly aligned with the x or y axis
if x_max == x_min: x_max = x_min + 1
if y_max == y_min: y_max = y_min + 1
# Renormalize to height [0,10000] keeping the aspect ratio
H = 10000.0
W = H * (x_max - x_min) / (y_max - y_min)
for trace_key, trace_v in traces.iteritems():
trace_v.calc_coords_h10000(H, W, x_min, y_min, x_max, y_max)
return traces
def reduce_to_n_strokes(self, n_strokes):
"""
Combines successive strokes based on minimum distance criteria
until reaching n_strokes, the desired number of strokes
"""
if n_strokes == len(self.strokes):
return
elif n_strokes <= 0:
print("Invalid number of strokes")
elif n_strokes>len(self.strokes):
print("Cannot reduce to {} strokes as {} > to the number of strokes ({})".format(n,n,len(self.strokes)))
else:
#Distance between the end point of a stroke and the start of the following one
dists = [dist(s1.get_x()[-1],s1.get_y()[-1],s2.get_x()[0],s2.get_y()[0])
for (s1, s2) in zip(self.strokes[:-1], self.strokes[1:])]
n_strokes_to_remove = len(self.strokes)-n_strokes
for i in range(n_strokes_to_remove):
idx_min = dists.index(min(dists))
x_pts = list(self.strokes[idx_min].get_x())+list(self.strokes[idx_min+1].get_x())
y_pts = list(self.strokes[idx_min].get_y())+list(self.strokes[idx_min+1].get_y())
self.strokes[idx_min] = Stroke(x_pts, y_pts)# Put the combination at min index
self.strokes.remove(self.strokes[idx_min+1])# Remove the second part of the new stroke from letter list
dists.remove(min(dists)) # Recompute the dists
def fromGML(gml):
"""Given a GML string, returns a tag object containing the GML's strokes."""
tag = Tag()
tag.gml = gml
root = etree.XML(gml)
strokes = []
for stroke in root.findall(".//stroke"):
points = []
for point in stroke.findall("pt"):
x = float(point.find('x').text)
y = float(point.find('y').text)
t = point.find('t')
t = float(t.text) if t is not None else 0
points.append((x, y, t))
strokes.append(Stroke(*points))
tag.strokes = strokes
return tag
import argparse
parser = argparse.ArgumentParser(description='study of log from ABC writer')
parser.add_argument('session', action="store",
help='pre-test_post-test analysis')
optimal_5 = "[0.384057971014 0.261567464485 0.197667629695 0.0816660085652 -0.00440566609228 -0.140472676258 -0.256491462802 -0.338229883296 -0.382085487718 -0.413521090541 -0.413038791269 -0.416852974564 -0.420292715375 -0.420952599295 -0.412451838673 -0.420285420695 -0.428258583881 -0.434183083379 -0.442016586753 -0.442028974646 -0.442042581755 -0.434874109522 -0.434786078475 -0.420289855072 -0.377177061268 -0.348832691991 -0.291577806642 -0.254559216915 -0.172181549423 -0.105270602782 -0.0403056142805 0.0117410496839 0.0702649407217 0.117015203781 0.159713997184 0.206322013387 0.263973865091 0.30986560194 0.350400167764 0.387222072047 0.422061663579 0.443227255209 0.464479438812 0.485345200072 0.492747037646 0.492749510342 0.5 0.492697041196 0.486207150786 0.463432582702 0.448889052645 0.423030378814 0.405872095099 0.359220307334 0.309229885035 0.252660132608 0.202630777478 0.148939216217 0.0474020470811 -0.0102019550865 -0.0756732227515 -0.15100759168 -0.220039666369 -0.262536071943 -0.328317137166 -0.37047801217 -0.42196661685 -0.463476086258 -0.478361672553 -0.5 -0.760812885288 -0.760812885288 -0.760812885288 -0.760812885288 -0.760812884914 -0.760813193689 -0.760569410568 -0.76064966956 -0.761591995733 -0.750426566518 -0.706004483632 -0.670061295046 -0.626167738735 -0.590804732411 -0.540996823739 -0.473249427541 -0.408888761791 -0.336911619956 -0.257075397024 -0.208837548122 -0.172908506258 -0.128950296999 -0.0865850603221 -0.0579143345632 -0.0582550576214 -0.0651440274355 -0.0723563293764 -0.0724071806674 -0.0724104757739 -0.0668064786106 -0.0651585252653 -0.0649565859506 -0.0533215255616 -0.0514341862059 -0.0434773991235 -0.0255062918735 0.0111018342626 0.0317652884402 0.0511906315574 0.0804522584393 0.094235381725 0.10921646595 0.132584146751 0.174684833436 0.218095478309 0.268390017243 0.31889725964 0.354800925309 0.397668489941 0.462208572201 0.513195122134 0.562288998383 0.599754700718 0.655206019842 0.701549779901 0.722662160754 0.744024752732 0.753488785838 0.757116044753 0.761591995733 0.753729297504 0.736669429131 0.709311762987 0.694124562448 0.674334335046 0.650838258014 0.607224624956 0.55737555518 0.52131242334 0.478317549495]"
optimal_8 = "[-0.0243213471502 0.0904459346802 0.191879977758 0.247243028688 0.312891090082 0.368296981283 0.412849361758 0.461527515922 0.490121537675 0.5 0.489490843252 0.453843360683 0.379630108469 0.303186796174 0.189037555525 0.0791228466465 -0.00776954936266 -0.0805627647689 -0.166169011807 -0.238016595293 -0.319994071421 -0.359383062304 -0.423849299916 -0.452531343391 -0.462020679872 -0.462736824738 -0.442512346492 -0.423525114127 -0.395364769417 -0.350865228925 -0.239046322262 -0.128424743986 -0.028370511862 0.0736934132185 0.134984725525 0.208510014058 0.264964236436 0.317522563737 0.336346085375 0.316846967057 0.284559505165 0.259770825994 0.21768485669 0.153803068304 0.125789464008 0.0727135855481 0.0168916929503 -0.0287440210235 -0.0976567128322 -0.157609520968 -0.213648261006 -0.250113262377 -0.302327265797 -0.381254479447 -0.434108906693 -0.470637858234 -0.489273797899 -0.5 -0.489550546621 -0.480976641699 -0.462761050442 -0.443353281481 -0.41300996464 -0.375293124575 -0.339097749419 -0.294286429658 -0.226592759017 -0.183043755797 -0.173437914232 -0.127793658783 -1.0562031225 -1.04820360574 -1.02836322248 -1.00095349005 -0.980599372324 -0.956094845959 -0.912745691564 -0.857278166152 -0.792776256007 -0.701462207942 -0.565969457 -0.448209749574 -0.306196128675 -0.203167676581 -0.0863271025033 -0.032994443288 0.0333441306285 0.0993569872153 0.15461469303 0.236364888641 0.318239019832 0.376858546444 0.472278036731 0.572313201034 0.683344750067 0.784175984946 0.855493567219 0.93888438073 0.99590142047 1.02147806244 1.06707739146 1.08487094069 1.08597731816 1.07613883833 1.06473326067 1.02900967065 0.972948760234 0.895052363327 0.823693087116 0.713764533598 0.553867498814 0.439134254025 0.330490462435 0.212379669088 0.166851351856 0.0788993968472 0.00747744375169 -0.0385102414128 -0.0838168546241 -0.129655801401 -0.195230035259 -0.240090048148 -0.30575238173 -0.380587637759 -0.445232068228 -0.528384536461 -0.611984487128 -0.672269078703 -0.765406796366 -0.835994740504 -0.881867452356 -0.937868708212 -0.962912873264 -0.989929644196 -1.00586739405 -1.03404106579 -1.04845452808 -1.05715979346 -1.08597731816 -1.07607518728]"
optimal_h = "[-0.465644654167 -0.406814716731 -0.317143484748 -0.25655215878 -0.184507813269 -0.118441201034 -0.0553834723074 -0.00932181178393 0.0194183811841 0.0234560320212 0.0229741558399 0.0198558402606 0.00518230513048 -0.00561355548688 -0.023476604695 -0.0456369309396 -0.0609669212641 -0.0783039569565 -0.0958308290006 -0.107504694821 -0.122015898631 -0.136546908227 -0.141697091876 -0.152690297875 -0.160242702095 -0.167886986693 -0.175414043936 -0.181501300613 -0.186817315464 -0.202738509777 -0.205912789304 -0.210021834416 -0.213345106856 -0.218603215225 -0.221366744246 -0.221622582653 -0.222464888308 -0.217520769696 -0.213860829694 -0.209908890338 -0.209877010992 -0.199882227334 -0.177189355294 -0.159481907181 -0.133505492852 -0.115017680319 -0.0931300084864 -0.0707479700553 -0.0477553518672 -0.0226805478612 -0.000729895945079 0.0172570655237 0.0322046529855 0.0461909624767 0.0534823572588 0.0685681214828 0.086925268026 0.0966687412317 0.111186442505 0.137582010585 0.155940148501 0.18842436246 0.22434651268 0.276803976787 0.332196893204 0.372666728477 0.410512178739 0.442746638623 0.454999833182 0.465644654167 0.206077769288 0.179962153339 0.117240601299 0.0696779201395 0.00106180971476 -0.0685579308667 -0.150194344883 -0.220277212671 -0.281668917342 -0.333703471297 -0.400495776588 -0.449510260463 -0.477053243392 -0.491956223732 -0.5 -0.496146846969 -0.481924725827 -0.463591255741 -0.436289465091 -0.407659440751 -0.35158958548 -0.299581285651 -0.251533617647 -0.195213985023 -0.128588095061 -0.0693932595753 -0.00639604632691 0.0438122242381 0.0952219824064 0.164920079353 0.224070719923 0.278726506941 0.337625369441 0.403524200395 0.454650607261 0.480701551145 0.485719172567 0.46424038153 0.433112014726 0.404273230851 0.371250460824 0.323485169112 0.267361860776 0.222522910558 0.198115385606 0.183044438119 0.167694094733 0.167910095626 0.167967785496 0.171848896587 0.190358226595 0.220416453598 0.250870059972 0.284195782296 0.3245905463 0.364946574833 0.401887892924 0.441108532437 0.469566505624 0.492441707905 0.49993204743 0.5 0.492162200469 0.475168963272 0.423450794606 0.369193965061 0.314342686513 0.261903197016 0.231693565608 0.213711288209]"
ref_shape5 = literal_eval(optimal_5.replace(' ',', '))
ref_shape5 = np.reshape(ref_shape5, (-1, 1))
ref_stroke5 = Stroke()
ref_stroke5.stroke_from_xxyy(np.reshape(ref_shape5,len(ref_shape5)))
ref_stroke5.uniformize()
ref_stroke5.normalize()
ref_shape8 = literal_eval(optimal_8.replace(' ',', '))
ref_shape8 = np.reshape(ref_shape8, (-1, 1))
ref_stroke8 = Stroke()
ref_stroke8.stroke_from_xxyy(np.reshape(ref_shape8,len(ref_shape8)))
ref_stroke8.uniformize()
ref_stroke8.normalize()
ref_shapeh = literal_eval(optimal_h.replace(' ',', '))
ref_shapeh = np.reshape(ref_shapeh, (-1, 1))
ref_strokeh = Stroke()
ref_strokeh.stroke_from_xxyy(np.reshape(ref_shapeh,len(ref_shapeh)))
def respondToGoodDemonstration_modulo_phase(self, shape):
"""
will learn the demo only if it is close to the goal shape, but no matter the way it was drawn
return a boolean that says if the stroke was accepted or not
"""
"""
Algo:
1) takes the shape of the demonstration and check if it is good enought to be learned
2) takes the curent learned shape
3) re-performs PCA taking in account the domonstrated shape,
then obtains a new space with new eigen vectors
4) projects demonstrated and learned shapes into this new space
and gets their new parameters
5) updates the learned parameters as the algebric middle
between demonstrated parameters and curent learned parameters.
"""
demo_shape = ShapeModeler.normaliseShapeWidth(numpy.array(shape))
reference = self.shapeModeler.getReference()
# create stroke from shape :
demo_stroke = Stroke()
demo_stroke.stroke_from_xxyy(numpy.reshape(demo_shape,len(demo_shape)))
demo_stroke.uniformize()
# create stroke from reference :
ref_stroke = Stroke()
ref_stroke.stroke_from_xxyy(reference)
ref_stroke.uniformize()
# get distance between demo and ref, modulo way and phase :
new_x,new_y,_,dist,_,_ = stroke.best_aligment(demo_stroke,ref_stroke)
demo_shape = numpy.concatenate((new_x,new_y),axis=0)
demo_shape = numpy.reshape(demo_shape, (-1, 1))
accepted = True
if dist>0.4:
accepted = False
# take the shape corresponding to the curent learned parameters in the curent space
learned_shape = self.shapeModeler.makeShape(self.params)
# if good shape, add the demo shape to the matrix for PCA and re-compute reference-shape params
if accepted:
self.shapeModeler.extendDataMat(demo_shape)
ref_params = self.shapeModeler.refParams
# re-compute parameters of the learned shape and the demo shape in the new PCA-space
params_demo, _ = self.shapeModeler.decomposeShape(demo_shape)
self.params, _ = self.shapeModeler.decomposeShape(learned_shape)
#d to get distance with clusters
# learning, if good shape :
diff_params = params_demo - self.params
if accepted:
#self.params = params_demo
self.params += diff_params/2 #go towards the demonstrated shape
# check new generated shape :
new_shape = self.shapeModeler.makeShape(self.params)
new_stroke = Stroke()
new_stroke.stroke_from_xxyy(numpy.reshape(demo_shape,len(new_shape)))
_,dist1 = stroke.euclidian_distance(new_stroke,ref_stroke)
if dist1>0.4:
self.params = params_demo
#self.params[self.paramsToVary[0]-1] = params_demo[self.paramsToVary[0]-1] #ONLY USE FIRST PARAM
#store it as an attempt (this isn't super appropriate but whatever)
'''if (self.doGroupwiseComparison):
newParamValue = self.params[
self.paramsToVary[0] - 1, 0] #USE ONLY FIRST PARAM FOR SELF-LEARNING ALGORITHM ATM
#print('Demo params: '+str(self.params))
bisect.insort(self.params_sorted, newParamValue)
self.shapeToParamsMapping.append(self.params)
#self.respondToFeedback(len(self.params_sorted)-3) # give feedback of most recent shape so bounds modify'''
return accepted, self.shapeModeler.makeShape(self.params), self.params, params_demo
ind = event.ind
for i in ind:
stroke2print = lookup_basedon_type_str_gain(round1, np.take(x, i),
np.take(y, i))
stroke2print.stroke_print()
if __name__ == '__main__':
#run only once on a round of data unless comment out save_round()
data_import = pd.read_csv(
'rounds\\Uploaded\\121418_LaurelCreek_Palmer.csv',
names=['Club', 'Loc Start', 'Dist Start'])
round1 = Round('12/14/18', 'LaurelCreek', 'Palmer', '18')
for idx, row in data_import.iterrows():
a = Stroke((idx + 1), row["Club"], row["Loc Start"], row["Dist Start"])
round1.add_stroke(a)
round1.calculate_strokes_gained()
round1.strokes = round1.strokes_to_pandas()
round1.calculate_strokes_to_hole()
round1.build_scorecard()
round1.single_round_plot()
#round1.save_round('rounds\Taillie_History.csv')
# fig2 = round1.plot_strgain_type_bar()
#
#
#
# full_disp = plt.figure()
#
def respondToGoodDemonstration_modulo_phase(self, shape):
"""
will learn the demo only if it is close to the goal shape, but no matter the way it was drawn
return a boolean that says if the stroke was accepted or not
"""
"""
Algo:
1) takes the shape of the demonstration and check if it is good enought to be learned
2) takes the curent learned shape
3) re-performs PCA taking in account the domonstrated shape,
then obtains a new space with new eigen vectors
4) projects demonstrated and learned shapes into this new space
and gets their new parameters
5) updates the learned parameters as the algebric middle
between demonstrated parameters and curent learned parameters.
"""
demo_shape = ShapeModeler.normaliseShapeWidth(numpy.array(shape))
reference = self.shapeModeler.getReference()
# create stroke from shape :
demo_stroke = Stroke()
demo_stroke.stroke_from_xxyy(numpy.reshape(demo_shape,
len(demo_shape)))
demo_stroke.uniformize()
# create stroke from reference :
ref_stroke = Stroke()
ref_stroke.stroke_from_xxyy(reference)
ref_stroke.uniformize()
# get distance between demo and ref, modulo way and phase :
new_x, new_y, _, dist, _, _ = stroke.best_aligment(
demo_stroke, ref_stroke)
demo_shape = numpy.concatenate((new_x, new_y), axis=0)
demo_shape = numpy.reshape(demo_shape, (-1, 1))
accepted = True
if dist > 0.4:
accepted = False
# take the shape corresponding to the curent learned parameters in the curent space
learned_shape = self.shapeModeler.makeShape(self.params)
# if good shape, add the demo shape to the matrix for PCA and re-compute reference-shape params
if accepted:
self.shapeModeler.extendDataMat(demo_shape)
ref_params = self.shapeModeler.refParams
# re-compute parameters of the learned shape and the demo shape in the new PCA-space
params_demo, _ = self.shapeModeler.decomposeShape(demo_shape)
self.params, _ = self.shapeModeler.decomposeShape(learned_shape)
#d to get distance with clusters
# learning, if good shape :
diff_params = params_demo - self.params
if accepted:
#self.params = params_demo
self.params += diff_params / 2 #go towards the demonstrated shape
# check new generated shape :
new_shape = self.shapeModeler.makeShape(self.params)
new_stroke = Stroke()
new_stroke.stroke_from_xxyy(
numpy.reshape(demo_shape, len(new_shape)))
_, dist1 = stroke.euclidian_distance(new_stroke, ref_stroke)
if dist1 > 0.4:
self.params = params_demo
#self.params[self.paramsToVary[0]-1] = params_demo[self.paramsToVary[0]-1] #ONLY USE FIRST PARAM
#store it as an attempt (this isn't super appropriate but whatever)
'''if (self.doGroupwiseComparison):
newParamValue = self.params[
self.paramsToVary[0] - 1, 0] #USE ONLY FIRST PARAM FOR SELF-LEARNING ALGORITHM ATM
#print('Demo params: '+str(self.params))
bisect.insort(self.params_sorted, newParamValue)
self.shapeToParamsMapping.append(self.params)
#self.respondToFeedback(len(self.params_sorted)-3) # give feedback of most recent shape so bounds modify'''
return accepted, self.shapeModeler.makeShape(
self.params), self.params, params_demo
def flattened_stroke(self):
"""Returns a stroke containing of all tag's strokes concatenated together"""
return sum((stroke for stroke in self.strokes), Stroke())
import argparse
parser = argparse.ArgumentParser(description='study of log from ABC writer')
parser.add_argument('session',
action="store",
help='pre-test_post-test analysis')
optimal_5 = "[0.384057971014 0.261567464485 0.197667629695 0.0816660085652 -0.00440566609228 -0.140472676258 -0.256491462802 -0.338229883296 -0.382085487718 -0.413521090541 -0.413038791269 -0.416852974564 -0.420292715375 -0.420952599295 -0.412451838673 -0.420285420695 -0.428258583881 -0.434183083379 -0.442016586753 -0.442028974646 -0.442042581755 -0.434874109522 -0.434786078475 -0.420289855072 -0.377177061268 -0.348832691991 -0.291577806642 -0.254559216915 -0.172181549423 -0.105270602782 -0.0403056142805 0.0117410496839 0.0702649407217 0.117015203781 0.159713997184 0.206322013387 0.263973865091 0.30986560194 0.350400167764 0.387222072047 0.422061663579 0.443227255209 0.464479438812 0.485345200072 0.492747037646 0.492749510342 0.5 0.492697041196 0.486207150786 0.463432582702 0.448889052645 0.423030378814 0.405872095099 0.359220307334 0.309229885035 0.252660132608 0.202630777478 0.148939216217 0.0474020470811 -0.0102019550865 -0.0756732227515 -0.15100759168 -0.220039666369 -0.262536071943 -0.328317137166 -0.37047801217 -0.42196661685 -0.463476086258 -0.478361672553 -0.5 -0.760812885288 -0.760812885288 -0.760812885288 -0.760812885288 -0.760812884914 -0.760813193689 -0.760569410568 -0.76064966956 -0.761591995733 -0.750426566518 -0.706004483632 -0.670061295046 -0.626167738735 -0.590804732411 -0.540996823739 -0.473249427541 -0.408888761791 -0.336911619956 -0.257075397024 -0.208837548122 -0.172908506258 -0.128950296999 -0.0865850603221 -0.0579143345632 -0.0582550576214 -0.0651440274355 -0.0723563293764 -0.0724071806674 -0.0724104757739 -0.0668064786106 -0.0651585252653 -0.0649565859506 -0.0533215255616 -0.0514341862059 -0.0434773991235 -0.0255062918735 0.0111018342626 0.0317652884402 0.0511906315574 0.0804522584393 0.094235381725 0.10921646595 0.132584146751 0.174684833436 0.218095478309 0.268390017243 0.31889725964 0.354800925309 0.397668489941 0.462208572201 0.513195122134 0.562288998383 0.599754700718 0.655206019842 0.701549779901 0.722662160754 0.744024752732 0.753488785838 0.757116044753 0.761591995733 0.753729297504 0.736669429131 0.709311762987 0.694124562448 0.674334335046 0.650838258014 0.607224624956 0.55737555518 0.52131242334 0.478317549495]"
optimal_8 = "[-0.0243213471502 0.0904459346802 0.191879977758 0.247243028688 0.312891090082 0.368296981283 0.412849361758 0.461527515922 0.490121537675 0.5 0.489490843252 0.453843360683 0.379630108469 0.303186796174 0.189037555525 0.0791228466465 -0.00776954936266 -0.0805627647689 -0.166169011807 -0.238016595293 -0.319994071421 -0.359383062304 -0.423849299916 -0.452531343391 -0.462020679872 -0.462736824738 -0.442512346492 -0.423525114127 -0.395364769417 -0.350865228925 -0.239046322262 -0.128424743986 -0.028370511862 0.0736934132185 0.134984725525 0.208510014058 0.264964236436 0.317522563737 0.336346085375 0.316846967057 0.284559505165 0.259770825994 0.21768485669 0.153803068304 0.125789464008 0.0727135855481 0.0168916929503 -0.0287440210235 -0.0976567128322 -0.157609520968 -0.213648261006 -0.250113262377 -0.302327265797 -0.381254479447 -0.434108906693 -0.470637858234 -0.489273797899 -0.5 -0.489550546621 -0.480976641699 -0.462761050442 -0.443353281481 -0.41300996464 -0.375293124575 -0.339097749419 -0.294286429658 -0.226592759017 -0.183043755797 -0.173437914232 -0.127793658783 -1.0562031225 -1.04820360574 -1.02836322248 -1.00095349005 -0.980599372324 -0.956094845959 -0.912745691564 -0.857278166152 -0.792776256007 -0.701462207942 -0.565969457 -0.448209749574 -0.306196128675 -0.203167676581 -0.0863271025033 -0.032994443288 0.0333441306285 0.0993569872153 0.15461469303 0.236364888641 0.318239019832 0.376858546444 0.472278036731 0.572313201034 0.683344750067 0.784175984946 0.855493567219 0.93888438073 0.99590142047 1.02147806244 1.06707739146 1.08487094069 1.08597731816 1.07613883833 1.06473326067 1.02900967065 0.972948760234 0.895052363327 0.823693087116 0.713764533598 0.553867498814 0.439134254025 0.330490462435 0.212379669088 0.166851351856 0.0788993968472 0.00747744375169 -0.0385102414128 -0.0838168546241 -0.129655801401 -0.195230035259 -0.240090048148 -0.30575238173 -0.380587637759 -0.445232068228 -0.528384536461 -0.611984487128 -0.672269078703 -0.765406796366 -0.835994740504 -0.881867452356 -0.937868708212 -0.962912873264 -0.989929644196 -1.00586739405 -1.03404106579 -1.04845452808 -1.05715979346 -1.08597731816 -1.07607518728]"
optimal_h = "[-0.465644654167 -0.406814716731 -0.317143484748 -0.25655215878 -0.184507813269 -0.118441201034 -0.0553834723074 -0.00932181178393 0.0194183811841 0.0234560320212 0.0229741558399 0.0198558402606 0.00518230513048 -0.00561355548688 -0.023476604695 -0.0456369309396 -0.0609669212641 -0.0783039569565 -0.0958308290006 -0.107504694821 -0.122015898631 -0.136546908227 -0.141697091876 -0.152690297875 -0.160242702095 -0.167886986693 -0.175414043936 -0.181501300613 -0.186817315464 -0.202738509777 -0.205912789304 -0.210021834416 -0.213345106856 -0.218603215225 -0.221366744246 -0.221622582653 -0.222464888308 -0.217520769696 -0.213860829694 -0.209908890338 -0.209877010992 -0.199882227334 -0.177189355294 -0.159481907181 -0.133505492852 -0.115017680319 -0.0931300084864 -0.0707479700553 -0.0477553518672 -0.0226805478612 -0.000729895945079 0.0172570655237 0.0322046529855 0.0461909624767 0.0534823572588 0.0685681214828 0.086925268026 0.0966687412317 0.111186442505 0.137582010585 0.155940148501 0.18842436246 0.22434651268 0.276803976787 0.332196893204 0.372666728477 0.410512178739 0.442746638623 0.454999833182 0.465644654167 0.206077769288 0.179962153339 0.117240601299 0.0696779201395 0.00106180971476 -0.0685579308667 -0.150194344883 -0.220277212671 -0.281668917342 -0.333703471297 -0.400495776588 -0.449510260463 -0.477053243392 -0.491956223732 -0.5 -0.496146846969 -0.481924725827 -0.463591255741 -0.436289465091 -0.407659440751 -0.35158958548 -0.299581285651 -0.251533617647 -0.195213985023 -0.128588095061 -0.0693932595753 -0.00639604632691 0.0438122242381 0.0952219824064 0.164920079353 0.224070719923 0.278726506941 0.337625369441 0.403524200395 0.454650607261 0.480701551145 0.485719172567 0.46424038153 0.433112014726 0.404273230851 0.371250460824 0.323485169112 0.267361860776 0.222522910558 0.198115385606 0.183044438119 0.167694094733 0.167910095626 0.167967785496 0.171848896587 0.190358226595 0.220416453598 0.250870059972 0.284195782296 0.3245905463 0.364946574833 0.401887892924 0.441108532437 0.469566505624 0.492441707905 0.49993204743 0.5 0.492162200469 0.475168963272 0.423450794606 0.369193965061 0.314342686513 0.261903197016 0.231693565608 0.213711288209]"
ref_shape5 = literal_eval(optimal_5.replace(' ', ', '))
ref_shape5 = np.reshape(ref_shape5, (-1, 1))
ref_stroke5 = Stroke()
ref_stroke5.stroke_from_xxyy(np.reshape(ref_shape5, len(ref_shape5)))
ref_stroke5.uniformize()
ref_stroke5.normalize()
ref_shape8 = literal_eval(optimal_8.replace(' ', ', '))
ref_shape8 = np.reshape(ref_shape8, (-1, 1))
ref_stroke8 = Stroke()
ref_stroke8.stroke_from_xxyy(np.reshape(ref_shape8, len(ref_shape8)))
ref_stroke8.uniformize()
ref_stroke8.normalize()
ref_shapeh = literal_eval(optimal_h.replace(' ', ', '))
ref_shapeh = np.reshape(ref_shapeh, (-1, 1))
ref_strokeh = Stroke()
ref_strokeh.stroke_from_xxyy(np.reshape(ref_shapeh, len(ref_shapeh)))