def experiment(transactions, epsilon, delta, mu, supp, rep):
results = []
for r in xrange(rep):
ts = time.time()
# Compute the bounds
t_bound = toivonen_bound(epsilon, delta)
d_bound = dbound_bound(len(transactions), epsilon, delta,
float(d_index), 0.5)
print(
'[+] Performing experiment {}/{}. (epsilon: {}, delta: {}, mu: {}, support: {})'
.format(r + 1, rep, epsilon, delta, mu, supp))
# Draw a sample --- Toivonen's bound
sample = sample_n(transactions, t_bound)
threshold = toivonen_threshold(len(sample), mu)
print('[|] Toivonen\'s bound sample size: {:,} ({})'.format(
len(sample), threshold))
t_fi = fim(sample, supp=supp, zmin=2)
t_fi_size = len(t_fi)
print('[|] Frequent itemsets found {}'.format(t_fi_size))
# Draw a sample --- d-bound
sample = sample_n(transactions, d_bound)
print('[|] d-bound sample size: {:,}'.format(len(sample)))
d_fi = fim(sample, supp=supp, zmin=2)
d_fi_size = len(d_fi)
print('[|] Frequent itemsets found {}'.format(d_fi_size))
print('[+] Comparing the frequent itemsets')
fi_common = compare_fis(t_fi, d_fi)
fi_common_size = len(fi_common)
percent = numpy.mean([
percentage(t_fi_size, fi_common_size),
percentage(d_fi_size, fi_common_size)
])
print('[|] {} frequent itemsets in common ({:.2f}%)'.format(
fi_common_size, percent))
# Finish the experiment
te = time.time()
milliseconds = (te - ts) * 1000.0
print('[+] Experiment duration: {:.4f} ms'.format(milliseconds))
print('')
results.append([
epsilon, delta, mu, t_bound, t_fi_size, threshold, d_bound,
d_fi_size, fi_common_size, percent, milliseconds
])
return numpy.mean(results, axis=0).tolist()
def _get_frequent_candidates(transactions, min_support, max_words):
"""
Runs an frequent itemset mining algorithms on the transactions
Input:
transactions: list of tuples representing items on each transaction
min_support: minimum support of an item set (positive: percentage,
negative: absolute number)
max_words: maximum number of items in each frequent set
Output:
list of extracted frequent itemset in the form (itemsets, support)
"""
return fim.fim(transactions, supp=min_support, zmax=max_words)
# print(test_pids_cat2[0])
# print(test_known_tracks[test_pids_cat2[0]])
# print([x[1] for x in test_known_tracks[test_pids_cat2[0]]])
#
# print("start")
sequences = urm_to_sequences(urm_pos=dr.get_position_matrix(position_type='last'),
target_list=[x[1] for x in test_known_tracks[test_pids_cat2[0]]],
min_common=1)
# for s in sequences: print(s)
# for s in sequences[0:2]:
# print("seuences:", s)
# print("maximal")
seq = fim(sequences[0:2], target='maximal', supp=-2, zmin=2, report='a')
# for s in seq:
# print("max>", s)
# print("normale")
# seq = fim(sequences[0:10], supp=-2, zmin=2, report='a')
# print("norm", seq)
# print("prefixspan")
sequences_for_prefix = urm_to_sequences(urm_pos=dr.get_position_matrix(position_type='last'),
target_list=[x[1] for x in test_known_tracks[test_pids_cat2[0]]],
min_common=1,
list_of_list_of_listss=True)
model = PrefixSpan.train(sequences_for_prefix, minSupport=0.1, maxPatternLength=250)
result = model.freqSequences().collect()
# --- save generated pattern spectrum ---
if cnt > 0 and pspfn != "": # if file name for pattern spectrum
t = time() # start timer, print log message
stderr.write('writing %s ... ' % pspfn)
with open(pspfn, 'w') as out: # write pattern spectrum to a file
for s in sorted([(z,c,psp[z,c]) for z,c in psp]):
out.write(('%d'+pssep+'%d'+pssep+'%.16g\n') % s)
stderr.write('[%d signature(s)]' % len(psp))
stderr.write(' done [%.2fs].\n' % (time()-t))
# --- analyze original data set ---
if len(args) < 2: exit() # check for an output file name
t = time() # start timer, print log message
stderr.write('analyzing original data ... ')
pats = fim(tracts, target, supp, zmin, zmax, 'a', border=border)
stderr.write('[%d pattern(s)]' % len(pats))
stderr.write(' done [%.2fs].\n' % (time()-t))
# --- pattern set reduction ---
if pred != 'x': # if to filter with pattern spectrum
t = time() # start timer, print log message
stderr.write('reducing pattern set ... ')
pats = patred(pats, pred, border, False)
stderr.write('[%d pattern(s)]' % len(pats))
stderr.write(' done [%.2fs].\n' % (time()-t))
# --- write output file ---
t = time() # start timer, print log message
stderr.write('writing %s ... ' % args[1])
with open(args[1], 'w') as out:
tid = int(argv[1])
if tid < -2:
print(fpgrowth.__doc__)
elif tid < -1:
print(eclat.__doc__)
elif tid < 0:
print(apriori.__doc__)
else:
tracts = [[1, 2, 3], [1, 4, 5], [2, 3, 4], [1, 2, 3, 4], [2, 3], [1, 2, 4],
[4, 5], [1, 2, 3, 4], [3, 4, 5], [1, 2, 3]]
print('transactions:')
for t in tracts:
print(t)
if tid < 1:
print('apriori(tracts, supp=-3, zmin=2):')
for r in apriori(tracts, supp=-3, zmin=2):
print r
elif tid < 2:
print('eclat(tracts, supp=-3, zmin=2):')
for r in eclat(tracts, supp=-3, zmin=2):
print r
elif tid < 3:
print('fpgrowth(tracts, supp=-3, zmin=2):')
for r in fpgrowth(tracts, supp=-3, zmin=2):
print r
else:
print('fim(tracts, supp=-3, zmin=2, report=\'#\'):')
for r in fim(tracts, supp=-3, zmin=2, report='#'):
print r
print(fpgrowth.__doc__)
elif tid < -1:
print(eclat.__doc__)
elif tid < 0:
print(apriori.__doc__)
else:
tracts = [ [ 1, 2, 3 ],
[ 1, 4, 5 ],
[ 2, 3, 4 ],
[ 1, 2, 3, 4 ],
[ 2, 3 ],
[ 1, 2, 4 ],
[ 4, 5 ],
[ 1, 2, 3, 4 ],
[ 3, 4, 5 ],
[ 1, 2, 3 ] ]
print('transactions:')
for t in tracts: print(t)
if tid < 1:
print ('apriori(tracts, supp=-3, zmin=2):')
for r in apriori(tracts, supp=-3, zmin=2): print r
elif tid < 2:
print ('eclat(tracts, supp=-3, zmin=2):')
for r in eclat(tracts, supp=-3, zmin=2): print r
elif tid < 3:
print ('fpgrowth(tracts, supp=-3, zmin=2):')
for r in fpgrowth(tracts, supp=-3, zmin=2): print r
else:
print ('fim(tracts, supp=-3, zmin=2, report=\'#\'):')
for r in fim(tracts, supp=-3, zmin=2, report='#'): print r
#human initial state
I0 = data_case[0] / abs(opt_repH)
E0 = 2 * data_case[0] / abs(opt_repH)
S0 = 1.0 - I0 - E0
#mosquito initial state
A0 = 1.0
S0_m = 1.0 / abs(opt_mu_m)
E0_m = 0.0
I0_m = 0.0
ini = [S0, E0, I0, A0, S0_m, E0_m, I0_m]
time_step = np.array(range(0, 7 * (53 - sim_start), 7))
#simulated case data
dengue = dengue_model()
opt_res = dengue.ode_run(dengue.model, ini, time_step, param)
res_inc_t1 = np.append(np.array([0]),
spi.cumtrapz(abs(opt_repH) * alpha * opt_res[:, 1]))
res_inc_t0 = np.append(np.array([0]),
spi.cumtrapz(abs(opt_repH) * alpha * opt_res[0:-1, 1]))
data_mw = 7 * (res_inc_t1 + 6.0 / 7.0) - 7 * np.append(np.array(
[0]), (res_inc_t0 + 6.0 / 7.0))
#print 'data_mw', data_mw
#fisher information matrix
fim_dengue = fim(dengue.model, time_step)
fim = fim_dengue.Fim_x(param, data_mw) #default pert = 0.01
fim_r = fim_dengue.Fim_rank(fim) #set the default tolerance to match matlab
print fim
print 'rank: ', fim_r
#-----------------------------------------------------------------------
if __name__ == '__main__':
runs = int(argv[1]) if len(argv) > 1 else 1
tracts = [[i+1 for i in range(100) if random() < 0.1]
for k in range(1000)]
with open('data.txt', 'w') as out:
for t in tracts:
for i in t: out.write('%d ' % i)
out.write('\n')
stderr.write('frequent item sets:\n')
stderr.write('fim ... '); t = time()
for r in range(runs):
pypats = fim(tracts, supp=-2, zmin=2, report='a')
stderr.write('done [%.3fs].\n' % (time()-t))
ref = set([(tuple(sorted(list(s))), x) for s,x in pypats])
stderr.write('\n')
for p,f in [('apriori', apriori),
('eclat', eclat),
('fpgrowth', fpgrowth),
('sam', sam),
('relim', relim)]:
stderr.write(p +' (all)\n')
stderr.write('python ... '); t = time()
for r in range(runs):
pypats = f(tracts, supp=-2, zmin=2, report='a')
stderr.write('done [%.3fs].\n' % (time()-t))
