def proficiency2(self):
"weighted average of n^2 proficiencies by entropy*entropy"
top = 0
ae = dict([(c, util.bin_entropy(self.observations, n))
for (c, n) in self.actual.iteritems()])
spe = 0 # sum of predicted entropies
for (pc, pn) in self.predicted.iteritems():
pe = util.bin_entropy(self.observations, pn)
spe += pe
if pe > 0:
for (ac, an) in self.actual.iteritems():
top += ae[ac] * pe * util.bin_mutual_info(
self.observations, pn, an, self.tp.get((ac, pc), 0))
return top / (spe * sum(ae.itervalues()))
def proficiency2 (self):
"weighted average of n^2 proficiencies by entropy*entropy"
top = 0
ae = dict([(c,util.bin_entropy(self.observations,n))
for (c,n) in self.actual.iteritems()])
spe = 0 # sum of predicted entropies
for (pc,pn) in self.predicted.iteritems():
pe = util.bin_entropy(self.observations,pn)
spe += pe
if pe > 0:
for (ac,an) in self.actual.iteritems():
top += ae[ac] * pe * util.bin_mutual_info(
self.observations,pn,an,self.tp.get((ac,pc),0))
return top / (spe * sum(ae.itervalues()))
def proficiency_raw (self):
mutual_information = 0.0
actual_entropy = 0.0
for (c,an) in self.actual.iteritems():
ae = util.bin_entropy(self.observations,an)
actual_entropy += ae
mutual_information += util.bin_mutual_info(
self.observations,self.predicted.get(c,0),an,self.tp.get((c,c),0))
if actual_entropy == 0:
return 0
return mutual_information / actual_entropy
def proficiency_raw(self):
mutual_information = 0.0
actual_entropy = 0.0
for (c, an) in self.actual.iteritems():
ae = util.bin_entropy(self.observations, an)
actual_entropy += ae
mutual_information += util.bin_mutual_info(
self.observations, self.predicted.get(c, 0), an,
self.tp.get((c, c), 0))
if actual_entropy == 0:
return 0
return mutual_information / actual_entropy
def assignment(self):
# https://pypi.python.org/pypi/munkres
import munkres
taxonomy = list(self.check_taxonomy())
costs = []
actual_entropy = []
for ac in taxonomy:
an = self.actual.get(ac, 0)
ae = util.bin_entropy(self.observations, an)
actual_entropy.append(ae)
if ae == 0:
costs.append([0 for _pc in taxonomy])
else:
# negative MI because munkres minimizes costs
costs.append([
-util.bin_mutual_info(self.observations,
self.predicted.get(pc, 0), an,
self.tp.get((ac, pc), 0))
for pc in taxonomy
])
m = munkres.Munkres()
indexes = m.compute(costs)
mutual_information = 0
reassigned = []
for row, col in indexes:
mutual_information += -costs[row][col]
if row != col:
ac = taxonomy[row]
pc = taxonomy[col]
c = -100 * costs[row][col]
if c > 0:
c /= actual_entropy[row]
reassigned.append(
(c, ac, self.actual.get(ac,
0), pc, self.predicted.get(pc, 0)))
if len(reassigned) > 0:
reassigned.sort(key=operator.itemgetter(0), reverse=True)
MuLabCat.logger.warn(
"Reassigned %d categories:\n%s", len(reassigned), "\n".join([
" Proficiency=%.2f%%: Actual [%s](%d) = Predicted [%s](%d)"
% (p, ac, an, pc, pn) for (p, ac, an, pc, pn) in reassigned
if (p >= 10 and an >= 5 and pn >= 5)
]))
actual_entropy = sum(actual_entropy)
return (taxonomy, indexes,
0 if actual_entropy == 0 else mutual_information /
actual_entropy)
def assignment (self):
# https://pypi.python.org/pypi/munkres
import munkres
taxonomy = list(self.check_taxonomy())
costs = []
actual_entropy = []
for ac in taxonomy:
an = self.actual.get(ac,0)
ae = util.bin_entropy(self.observations,an)
actual_entropy.append(ae)
if ae == 0:
costs.append([0 for _pc in taxonomy])
else:
# negative MI because munkres minimizes costs
costs.append([- util.bin_mutual_info(
self.observations,self.predicted.get(pc,0),
an,self.tp.get((ac,pc),0))
for pc in taxonomy])
m = munkres.Munkres()
indexes = m.compute(costs)
mutual_information = 0
reassigned = []
for row, col in indexes:
mutual_information += - costs[row][col]
if row != col:
ac = taxonomy[row]
pc = taxonomy[col]
c = -100*costs[row][col]
if c > 0:
c /= actual_entropy[row]
reassigned.append((c,ac,self.actual.get(ac,0),pc,self.predicted.get(pc,0)))
if len(reassigned) > 0:
reassigned.sort(key=operator.itemgetter(0),reverse=True)
MuLabCat.logger.warn("Reassigned %d categories:\n%s",len(reassigned),"\n".join(
[" Proficiency=%.2f%%: Actual [%s](%d) = Predicted [%s](%d)" % (p,ac,an,pc,pn)
for (p,ac,an,pc,pn) in reassigned if (p >= 10 and an >= 5 and pn >= 5)]))
actual_entropy = sum(actual_entropy)
return (taxonomy,indexes, 0 if actual_entropy == 0 else mutual_information / actual_entropy)
