def compute_statistics(self):
"""
Compute max, min, average, median, standard deviation and
number of elements.
"""
if not self.statistics:
self.data_sorted = self.data[:]
self.data_sorted.sort()
self._min = self.data_sorted[0]
self._max = self.data_sorted[-1]
self._median = self.data_sorted[int(len(self.data)/2)]
self._mean = pylib_basics.mean(self.data)
self._sd = pylib_basics.standard_deviation(self.data)
self.statistics = True
def compute_statistics(self):
"""
Compute max, min, average, median, standard deviation and
number of elements.
"""
if not self.statistics:
self.data_sorted = self.data[:]
self.data_sorted.sort()
self._min = self.data_sorted[0]
self._max = self.data_sorted[-1]
self._median = self.data_sorted[int(len(self.data) / 2)]
self._mean = pylib_basics.mean(self.data)
self._sd = pylib_basics.standard_deviation(self.data)
self.statistics = True
(succ, count) = tree.classify_set(i[1],pylib_basics.verbose())
succ_percent = float(succ)/count*100
print "Test: %4d out of %4d, %7.3f%%" % (succ, count, succ_percent)
te_results.append((succ, count,succ_percent))
tr_percent = (map(lambda x:x[2],tr_results))
te_percent = (map(lambda x:x[2],te_results))
rig = (map(lambda x:x[3],tr_results))
depths = (map(lambda x:x[4],tr_results))
size = (map(lambda x:x[5],tr_results))
leaves = (map(lambda x:x[6],tr_results))
print "%s Splits %2d RIGL %5.3f RIG %5.3f+/-%5.3f (%5.2f+/-%4.2f, %7.2f+/-%4.2f, %7.2f+/-%4.2f) Train: %7.3f+/-%7.3f%% Test: %7.3f+/-%7.3f%%" %\
(feature_selection,
max_split,
relgain_limit,
pylib_basics.mean(rig),
pylib_basics.standard_deviation(rig),
pylib_basics.mean(depths),
pylib_basics.standard_deviation(depths),
pylib_basics.mean(size),
pylib_basics.standard_deviation(size),
pylib_basics.mean(leaves),
pylib_basics.standard_deviation(leaves),
pylib_basics.mean(tr_percent),
pylib_basics.standard_deviation(tr_percent),
pylib_basics.mean(te_percent),
pylib_basics.standard_deviation(te_percent))
else:
tree = dectree_constructor(set,
entropy_compare_fun,
(succ, count) = tree.classify_set(i[1],pylib_basics.verbose())
succ_percent = float(succ)/count*100
print "Test: %4d out of %4d, %7.3f%%" % (succ, count, succ_percent)
te_results.append((succ, count,succ_percent))
tr_percent = (map(lambda x:x[2],tr_results))
te_percent = (map(lambda x:x[2],te_results))
rig = (map(lambda x:x[3],tr_results))
depths = (map(lambda x:x[4],tr_results))
size = (map(lambda x:x[5],tr_results))
leaves = (map(lambda x:x[6],tr_results))
print "%s Splits %2d RIGL %5.3f RIG %5.3f+/-%5.3f (%5.2f+/-%4.2f, %7.2f+/-%4.2f, %7.2f+/-%4.2f) Train: %7.3f+/-%7.3f%% Test: %7.3f+/-%7.3f%%" %\
(feature_selection,
max_split,
relgain_limit,
pylib_basics.mean(rig),
pylib_basics.standard_deviation(rig),
pylib_basics.mean(depths),
pylib_basics.standard_deviation(depths),
pylib_basics.mean(size),
pylib_basics.standard_deviation(size),
pylib_basics.mean(leaves),
pylib_basics.standard_deviation(leaves),
pylib_basics.mean(tr_percent),
pylib_basics.standard_deviation(tr_percent),
pylib_basics.mean(te_percent),
pylib_basics.standard_deviation(te_percent))
else:
tree = dectree_constructor(set,
entropy_compare_fun,