def create_report_var(name, mean, cov, inf, minimum=None, maximum=None):
report = Node(id=name)
report.data("covariance", cov)
report.data("information", inf)
node_mean = report.data("mean", mean)
with node_mean.data_file("bounds", "image/png") as filename:
e = 3 * sqrt(cov.diagonal())
pylab.ioff()
f = pylab.figure()
x = range(len(mean))
pylab.errorbar(x, mean, yerr=e)
if minimum is not None:
pylab.plot(x, minimum, "b-")
if maximum is not None:
pylab.plot(x, maximum, "b-")
pylab.savefig(filename)
pylab.close(f)
f = report.figure(name)
f.sub("mean/bounds", caption="Mean")
f.sub("covariance", caption="Covariance", display="posneg")
f.sub("information", caption="Information", display="posneg")
return report
class Test(unittest.TestCase):
def setUp(self):
# n1 -- n2 -- n3 -- n4
# \_ n3bis
self.n1 = Node("n1")
self.n3bis = self.n1.data("n3", None)
self.n2 = self.n1.data("n2", None)
self.n3 = self.n2.data("n3", None)
self.n4 = self.n3.data("n4", None)
self.all = [self.n1, self.n2, self.n3, self.n4, self.n3bis]
def test_normal(self):
""" Testing normal cases of resolve_url """
self.assertEqual(self.n1, self.n2.resolve_url(".."))
self.assertEqual(self.n1, self.n3.resolve_url("../.."))
self.assertEqual(self.n1, self.n4.resolve_url("../../.."))
self.assertEqual(self.n2, self.n4.resolve_url("../../../n2"))
self.assertEqual(self.n2, self.n1.resolve_url("n2"))
self.assertEqual(self.n3, self.n1.resolve_url("n2/n3"))
def test_smart(self):
""" Testing smart cases of resolve_url """
self.assertEqual(self.n1.resolve_url("n3"), self.n3bis)
self.assertEqual(self.n2.resolve_url("n3"), self.n3)
self.assertEqual(self.n3.resolve_url("n3"), self.n3)
self.assertEqual(self.n2.resolve_url("n1/n3"), self.n3bis)
self.assertEqual(self.n3.resolve_url("n1/n3"), self.n3bis)
self.assertEqual(self.n4.resolve_url("n4"), self.n4)
self.assertEqual(self.n3.resolve_url("n4"), self.n4)
self.assertEqual(self.n2.resolve_url("n4"), self.n4)
self.assertEqual(self.n1.resolve_url("n4"), self.n4)
for n in self.all:
self.assertEqual(n.resolve_url("n4"), self.n4)
def test_relative(self):
for A in self.all:
for B in self.all:
if A == B:
continue
url = A.get_relative_url(B)
self.assertEqual(A.resolve_url_dumb(url), B)
def test_not_existent(self):
""" Testing that we warn if not existent """
self.assertRaises(NotExistent, self.n1.resolve_url, "x")
def test_not_found(self):
self.assertRaises(NotExistent, self.n1.resolve_url, "..")
def test_well_formed(self):
""" Testing that we can recognize invalid urls """
self.assertRaises(InvalidURL, self.n1.resolve_url, "")
self.assertRaises(InvalidURL, self.n2.resolve_url, "//")
class Test(unittest.TestCase):
def setUp(self):
# n1 -- n2 -- n3 -- n4
# \_ n3bis
self.n1 = Node("n1")
self.n3bis = self.n1.data("n3", None)
self.n2 = self.n1.data("n2", None)
self.n3 = self.n2.data("n3", None)
self.n4 = self.n3.data("n4", None)
self.all = [self.n1, self.n2, self.n3, self.n4, self.n3bis]
def test_normal(self):
""" Testing normal cases of resolve_url """
self.assertEqual(self.n1, self.n2.resolve_url(".."))
self.assertEqual(self.n1, self.n3.resolve_url("../.."))
self.assertEqual(self.n1, self.n4.resolve_url("../../.."))
self.assertEqual(self.n2, self.n4.resolve_url("../../../n2"))
self.assertEqual(self.n2, self.n1.resolve_url("n2"))
self.assertEqual(self.n3, self.n1.resolve_url("n2/n3"))
def test_smart(self):
""" Testing smart cases of resolve_url """
self.assertEqual(self.n1.resolve_url("n3"), self.n3bis)
self.assertEqual(self.n2.resolve_url("n3"), self.n3)
self.assertEqual(self.n3.resolve_url("n3"), self.n3)
self.assertEqual(self.n2.resolve_url("n1/n3"), self.n3bis)
self.assertEqual(self.n3.resolve_url("n1/n3"), self.n3bis)
self.assertEqual(self.n4.resolve_url("n4"), self.n4)
self.assertEqual(self.n3.resolve_url("n4"), self.n4)
self.assertEqual(self.n2.resolve_url("n4"), self.n4)
self.assertEqual(self.n1.resolve_url("n4"), self.n4)
for n in self.all:
self.assertEqual(n.resolve_url("n4"), self.n4)
def test_relative(self):
for A in self.all:
for B in self.all:
if A == B:
continue
url = A.get_relative_url(B)
self.assertEqual(A.resolve_url_dumb(url), B)
def test_not_existent(self):
""" Testing that we warn if not existent """
self.assertRaises(NotExistent, self.n1.resolve_url, "x")
def test_not_found(self):
self.assertRaises(NotExistent, self.n1.resolve_url, "..")
def test_well_formed(self):
""" Testing that we can recognize invalid urls """
self.assertRaises(InvalidURL, self.n1.resolve_url, "")
self.assertRaises(InvalidURL, self.n2.resolve_url, "//")
def testImage(self):
C = numpy.random.rand(50, 50)
information = pinv(C)
T = numpy.random.rand(50, 50, 3)
report = Node("rangefinder")
report.data("Tx", T[:, :, 0])
report.data("Ty", T[:, :, 1])
report.data("Tz", T[:, :, 2])
cov = report.data("covariance", C)
report.data("information", information)
pylab = get_pylab_instance()
with cov.data_file("plot", MIME_PNG) as f:
pylab.figure()
pylab.plot(C)
pylab.savefig(f)
pylab.close()
report.table("results", cols=["One", "Two"], data=[[1, 2], [3, 4]])
f = report.figure(caption="Covariance and information matrix", cols=3)
f.sub("covariance", "should default to plot")
f.sub("covariance/plot", "Same as <-")
f = report.figure("Tensors", cols=3)
f.sub("Tx", display="posneg")
f.sub("Ty", display="posneg")
f.sub("Tz", display="posneg")
self.node_serialization_ok(report)
def testImage(self):
C = numpy.random.rand(50, 50)
information = pinv(C)
T = numpy.random.rand(50, 50, 3)
report = Node('rangefinder')
report.data('Tx', T[:, :, 0])
report.data('Ty', T[:, :, 1])
report.data('Tz', T[:, :, 2])
cov = report.data('covariance', C)
report.data('information', information)
pylab = get_pylab_instance()
with cov.data_file('plot', 'image/png') as f:
pylab.figure()
pylab.plot(C)
pylab.savefig(f)
pylab.close()
report.table('results',
cols=['One', 'Two'],
data=[[1, 2], [3, 4]])
f = report.figure(caption='Covariance and information matrix', cols=3)
f.sub('covariance', 'should default to plot')
f.sub('covariance/plot', 'Same as <-')
f = report.figure('Tensors', cols=3)
f.sub('Tx', display='posneg')
f.sub('Ty', display='posneg')
f.sub('Tz', display='posneg')
self.node_serialization_ok(report)
def create_report_figure_tensors(T, report_id, same_scale=False, caption=None):
report = Node(report_id)
f = report.figure('tensors', caption=caption)
if same_scale:
max_value = abs(T).max()
else:
max_value = None
components = [(0, 'Tx', '$T_x$'),
(1, 'Ty', '$T_y$'),
(2, 'Ttheta', '$T_{\\theta}$')]
for component, slice, caption in components:
report.data(slice, T[component, :, :].squeeze())
f.sub(slice, display='posneg', max_value=max_value)
return report
def create_report_affine(state, report_id):
y_dot_mean = state.result.y_dot_stats.mean
y_dot_cov = state.result.y_dot_stats.covariance
y_dot_inf = state.result.y_dot_stats.information
y_dot_min = state.result.y_dot_stats.minimum
y_dot_max = state.result.y_dot_stats.maximum
y_mean = state.result.y_stats.mean
y_cov = state.result.y_stats.covariance
y_inf = state.result.y_stats.information
y_min = state.result.y_stats.minimum
y_max = state.result.y_stats.maximum
u_mean = state.result.u_stats.mean
u_inf = state.result.u_stats.information
u_cov = state.result.u_stats.covariance
N = state.result.N.get_value()
report_y_dot = create_report_var(
name="y_dot", mean=y_dot_mean, cov=y_dot_cov, inf=y_dot_inf, minimum=y_dot_min, maximum=y_dot_max
)
report_y = create_report_var(name="y", mean=y_mean, cov=y_cov, inf=y_inf, minimum=y_min, maximum=y_max)
report_u = create_report_var(name="u", mean=u_mean, cov=u_cov, inf=u_inf)
T = state.result.T.get_value()
report_T = create_report_figure_tensors(T, report_id="T", caption="Learned T")
Nreport = Node("n-report")
with Nreport.data("N", N).data_file("N", "image/png") as filename:
pylab.figure()
for i in range(N.shape[1]):
v = N[:, i].squeeze()
pylab.plot(v)
pylab.savefig(filename)
pylab.close()
f = Nreport.figure("nfig")
f.sub("N/N")
bT = state.result.bT.get_value()
report_bT = create_report_figure_tensors(bT, report_id="bT", caption="Learned bT")
bTn = state.result.bTn.get_value()
report_bTn = create_report_figure_tensors(bTn, report_id="bTn", caption="Learned bTn")
node = Node(id=report_id, children=[report_y, report_y_dot, report_u, report_T, Nreport, report_bT, report_bTn])
return node
def create_report_covariance(state, report_id):
covariance = state.result.cov_sensels
report = Node(report_id)
report.data('covariance', covariance)
report.data('correlation', cov2corr(covariance))
report.data('information', pinv(covariance, rcond=1e-2))
f = report.figure('matrices')
f.sub('covariance', caption='Covariance matrix', display='posneg')
f.sub('correlation', caption='Correlation matrix', display='posneg')
f.sub('information', caption='Information matrix', display='posneg')
return report
def main():
def find_in_path(path, pattern):
for root, dirs, files in os.walk(path): #@UnusedVariable
for f in files:
if fnmatch.fnmatch(f, pattern):
yield os.path.join(root, f)
dir = 'Bicocca_2009-02-25a/out/camera_bgds_predict/gray_GI_DI/'
files = list(find_in_path(dir, '*.??.pickle*'))
if not files:
raise Exception('No files found.')
for f in files:
print 'Loading {0}'.format(f)
data = my_pickle_load(f)
basename = os.path.splitext(f)[0]
out_html = basename + '/report.html'
#out_pickle = basename + '/report.pickle'
if not os.path.exists(basename):
os.makedirs(basename)
logdir = data['logdir']
time = data['time']
y = data['y']
y_dot = data['y_dot']
y_dot_pred = data['y_dot_pred']
y_dot_s = data['y_dot_s']
y_dot_pred_s = data['y_dot_pred_s']
report = Node('{logdir}-{time:.2f}'.format(logdir=logdir, time=time))
prod = -numpy.minimum(0, y_dot_s * y_dot_pred_s)
report.data('y', y)
report.data('y_dot', y_dot)
report.data('y_dot_pred', y_dot_pred)
report.data('y_dot_s', y_dot_s)
report.data('y_dot_pred_s', y_dot_pred_s)
report.data('prod', prod)
f = report.figure('display')
f.sub('y', display='rgb')
f.sub('y_dot', display='posneg')
f.sub('y_dot_pred', display='posneg')
f.sub('y_dot_s', display='posneg')
f.sub('y_dot_pred_s', display='posneg')
f.sub('prod', display='scale')
zones = [
('ZoneA', range(95, 181), range(55, 140)),
('ZoneB', range(80, 111), range(480, 510)),
('ZoneC', range(50, 181), range(330, 460))
]
for zone in zones:
name, y, x = zone
zr = report.node(name)
zf = zr.figure()
for var in ['y', 'y_dot', 'y_dot_pred', 'y_dot_s', 'y_dot_pred_s',
'prod']:
zd = data[var][y, :][:, x]
zr.data(var, zd)
if var == 'y':
disp = {'display': 'rgb'}
elif var == 'prod':
disp = {'display': 'scale'}
else:
disp = {'display': 'posneg'}
zf.sub(var, **disp)
outdir = basename + '/' + name
id = "%s:%.2f:%s" % (logdir, time, name)
print id
if not os.path.exists(outdir):
os.makedirs(outdir)
create_latex_frag(os.path.join(outdir, 'conf.tex'), outdir, zr, id, True)
create_latex_frag(os.path.join(outdir, 'report.tex'), outdir, zr, id, False)
print 'Writing on %s' % out_html
node_to_html_document(report, out_html)
def analyze_olfaction_covariance(covariance, receptors):
''' Covariance: n x n covariance matrix.
Positions: list of n positions '''
positions = [pose.get_2d_position() for pose, sens in receptors] #@UnusedVariable
positions = array(positions).transpose().squeeze()
require_shape(square_shape(), covariance)
n = covariance.shape[0]
require_shape((2, n), positions)
distances = create_distance_matrix(positions)
correlation = cov2corr(covariance)
flat_distances = distances.reshape(n * n)
flat_correlation = correlation.reshape(n * n)
# let's fit a polynomial
deg = 4
poly = polyfit(flat_distances, flat_correlation, deg=deg)
knots = linspace(min(flat_distances), max(flat_distances), 2000)
poly_int = polyval(poly, knots)
poly_fder = polyder(poly)
fder = polyval(poly_fder, distances)
Ttheta = create_olfaction_Ttheta(positions, fder)
Tx, Ty = create_olfaction_Txy(positions, fder, distances)
report = Node('olfaction-theory')
report.data('flat_distances', flat_distances)
report.data('flat_correlation', flat_correlation)
with report.data_file('dist_vs_corr', 'image/png') as filename:
pylab.figure()
pylab.plot(flat_distances, flat_correlation, '.')
pylab.plot(knots, poly_int, 'r-')
pylab.xlabel('distance')
pylab.ylabel('correlation')
pylab.title('Correlation vs distance')
pylab.legend(['data', 'interpolation deg = %s' % deg])
pylab.savefig(filename)
pylab.close()
with report.data('fder', fder).data_file('fder', 'image/png') as filename:
pylab.figure()
pylab.plot(knots, polyval(poly_fder, knots), 'r-')
pylab.title('f der')
pylab.savefig(filename)
pylab.close()
report.data('distances', distances)
report.data('correlation', correlation)
report.data('covariance', covariance)
report.data('f', polyval(poly, distances))
f = report.figure(id='cor-vs-distnace', caption='Estimated kernels',
shape=(3, 3))
f.sub('dist_vs_corr')
f.sub('fder')
f.sub('f', display='scale')
f.sub('distances', display='scale')
f.sub('correlation', display='posneg')
f.sub('covariance', display='posneg')
T = numpy.zeros(shape=(3, Tx.shape[0], Tx.shape[1]))
T[0, :, :] = Tx
T[1, :, :] = Ty
T[2, :, :] = Ttheta
T_report = create_report_figure_tensors(T, report_id='tensors',
caption="Predicted learned tensors")
report.add_child(T_report)
return report
