import os
from recsys.evaluation.ranking import AveragePrecision
from recsys.evaluation.ranking import ReciprocalRank
testfile = open("testset_txt_img_cat.list", "r")
resultsfolder = "../Java/Results/"
lines = testfile.xreadlines()
GT = []
ap = AveragePrecision()
rr = ReciprocalRank()
for line in lines:
GT.append(line.split()[2])
# GT.append(line.strip())
CCA = resultsfolder
for fname in os.listdir(CCA):
content = open(CCA + "/" + fname, "r")
values = content.xreadlines()
q = []
for each in values:
q.append(each.strip())
sum10 = 0.0
totmrrsum = 0
count = 0
count1 = 0
for i, j in zip(GT, q):
num = 0.0
denom = 1.0
localsum = 0.0
q1 = j.split()
for corr in q1:
if i == corr:
num = num + 1.0
def test_RANK_ReciprocalRank_add_entry(self):
rr= ReciprocalRank()
QUERY = 'invented'
rr.load(self.GT_DECISION, QUERY)
assert_equal(rr.compute(), 0.0)
def test_RANK_ReciprocalRank_add_entry(self):
rr = ReciprocalRank()
QUERY = 'invented'
rr.load(self.GT_DECISION, QUERY)
assert_equal(rr.compute(), 0.0)
def test_RANK_ReciprocalRank_compute(self):
rr = ReciprocalRank()
QUERY = 'instrumental'
assert_equal(rr.compute(self.GT_DECISION, QUERY), 0.5)
def test_RANK_ReciprocalRank_compute(self):
rr = ReciprocalRank()
QUERY = 'instrumental'
assert_equal(rr.compute(self.GT_DECISION, QUERY), 0.5)
