from pySVM import *
import sys

SV=SparseVector


v=SV(range(1, 10))



#
# classification
# 
cp = ClassificationProblem()
cp.reserve(100)

ce = ClassificationExample(v, 1)

if not  ce.label:
    sys.exit(1)
if v - ce.vector:
    sys.exit(1)

cp.append(ce)
cp.append(ClassificationExample(-v, 0))

if len(cp) != 2:
    sys.exit(1)

if cp[0].vector -ce.vector:
    sys.exit(1)

if cp[1].vector +v:
    sys.exit(1)

cp.normalize()

vv = cp[0].vector

sum = reduce(lambda x,y: x+y[1], vv, 0.0)

if int(sum) != 9:
    sys.exit(1)


sum = 0
ix = iter(cp)
del cp
for p in ix:
    sum += p.label

if sum-1.0:
    sys.exit(1)

#
# regression
# 
cp = RegressionProblem()
cp.reserve(100)

ce = RegressionExample(v, 1)

if not  ce.label:
    sys.exit(1)
if v - ce.vector:
    sys.exit(1)

cp.append(ce)
cp.append(RegressionExample(-v, 0))

if len(cp) != 2:
    sys.exit(1)

if cp[0].vector -ce.vector:
    sys.exit(1)

if cp[1].vector +v:
    sys.exit(1)

cp.normalize()

vv = cp[0].vector
sum = reduce(lambda x,y: x+y[1], vv, 0.0)

if int(sum) != 9:
    sys.exit(1)

sum = 0
ix = iter(cp)
del cp
for p in ix:
    sum += p.label

if sum-1.0:
    sys.exit(1)