def svd_protocol_evaluation(data_path, params):
solr = "http://localhost:8983/solr/grrecsys"
test_c = consumption(
ratings_path=data_path + 'test/test_N20.data',
rel_thresh=0,
with_ratings=True
) #debiera ser el test_c, pero como includeRated=False, da lo mismo
train_c = consumption(ratings_path=data_path + 'eval_train_N20.data',
rel_thresh=0,
with_ratings=False)
svd = pyreclab.SVD(dataset=data_path + 'eval_train_N20.data',
dlmchar=b',',
header=False,
usercol=0,
itemcol=1,
ratingcol=2)
svd.train(factors=params['f'],
maxiter=params['mi'],
lr=params['lr'],
lamb=params['lamb'])
recommendationList, map, ndcg = svd.testrec(input_file=data_path +
'test/test_N20.data',
dlmchar=b',',
header=False,
usercol=0,
itemcol=1,
ratingcol=2,
topn=100,
relevance_threshold=0,
includeRated=False)
MRRs = dict((N, []) for N in [5, 10, 15, 20])
nDCGs = dict((N, []) for N in [5, 10, 15, 20])
APs = dict((N, []) for N in [5, 10, 15, 20])
Rprecs = dict((N, []) for N in [5, 10, 15, 20])
for userId, recList in recommendationList.items():
book_recs = remove_consumed(
user_consumption=train_c[userId],
rec_list=recommendationList[userId]
) #da lo mismo este paso, según Gabriel el testrec no devuelve items consumidos
book_recs = recs_cleaner(solr=solr,
consumpt=train_c[userId],
recs=book_recs[:100])
recs = user_ranked_recs(user_recs=book_recs,
user_consumpt=test_c[userId])
for N in [5, 10, 15, 20]:
mini_recs = dict((k, recs[k]) for k in list(recs.keys())
[:N]) #python 3.x: list() es necesario
MRRs[N].append(MRR(recs=mini_recs, rel_thresh=1))
nDCGs[N].append(
nDCG(recs=mini_recs, alt_form=False, rel_thresh=False))
APs[N].append(AP_at_N(n=N, recs=recs, rel_thresh=1))
Rprecs[N].append(R_precision(n_relevants=N, recs=mini_recs))
for N in [5, 10, 15, 20]:
with open('TwitterRatings/funkSVD/clean/protocol.txt', 'a') as file:
file.write( "N=%s, nDCG=%s, MAP=%s, MRR=%s, R-precision=%s\n" % \
(N, mean(nDCGs[N]), mean(APs[N]), mean(MRRs[N]), mean(Rprecs[N])) )
def nDCGMAP_calculator(data_path, params, topN, output_filename):
user_consumption = consumption(ratings_path=data_path + 'ratings.total',
rel_thresh=0,
with_ratings=True)
svd = pyreclab.SVD(
dataset=data_path + 'ratings.train', #data_path+'train/train.'+str(i),
dlmchar=b',',
header=False,
usercol=0,
itemcol=1,
ratingcol=2)
svd.train(factors=params['f'],
maxiter=params['mi'],
lr=params['lr'],
lamb=params['lamb'])
recommendationList = svd.testrec(
input_file=data_path + 'test/' +
os.listdir(data_path + 'test/')[0], #data_path+'val/val.'+str(i),
dlmchar=b',',
header=False,
usercol=0,
itemcol=1,
ratingcol=2,
topn=100,
includeRated=False)
MRR_thresh4 = []
MRR_thresh3 = []
nDCGs_bin_thresh4 = dict((n, []) for n in topN)
nDCGs_bin_thresh3 = dict((n, []) for n in topN)
nDCGs_normal = dict((n, []) for n in topN)
nDCGs_altform = dict((n, []) for n in topN)
APs_thresh4 = dict((n, []) for n in topN)
APs_thresh3 = dict((n, []) for n in topN)
APs_thresh2 = dict((n, []) for n in topN)
for userId in recommendationList[0]:
recs = user_ranked_recs(user_recs=recommendationList[0][userId],
user_consumpt=user_consumption[userId])
MRR_thresh4.append(MRR(recs=recs, rel_thresh=4))
MRR_thresh3.append(MRR(recs=recs, rel_thresh=3))
for n in topN:
mini_recs = dict((k, recs[k]) for k in recs.keys()[:n])
nDCGs_bin_thresh4[n].append(
nDCG(recs=mini_recs, alt_form=False, rel_thresh=4))
nDCGs_bin_thresh3[n].append(
nDCG(recs=mini_recs, alt_form=False, rel_thresh=3))
nDCGs_normal[n].append(
nDCG(recs=mini_recs, alt_form=False, rel_thresh=False))
nDCGs_altform[n].append(
nDCG(recs=mini_recs, alt_form=True, rel_thresh=False))
APs_thresh4[n].append(AP_at_N(n=n, recs=recs, rel_thresh=4))
APs_thresh3[n].append(AP_at_N(n=n, recs=recs, rel_thresh=3))
APs_thresh2[n].append(AP_at_N(n=n, recs=recs, rel_thresh=2))
with open('TwitterRatings/funkSVD/' + output_filename, 'a') as file:
for n in topN:
file.write( "N=%s, normal nDCG=%s, alternative nDCG=%s, bin nDCG(rel_thresh=4)=%s, bin nDCG(rel_thresh=3)=%s, MAP(rel_thresh=4)=%s, MAP(rel_thresh=3)=%s, MAP(rel_thresh=2)=%s, MRR(rel_thresh=4)=%s, MRR(rel_thresh=3)=%s\n" % \
(n, mean(nDCGs_normal[n]), mean(nDCGs_altform[n]), mean(nDCGs_bin_thresh4[n]), mean(nDCGs_bin_thresh3[n]), mean(APs_thresh4[n]), mean(APs_thresh3[n]), mean(APs_thresh2[n]), mean(MRR_thresh4), mean(MRR_thresh3)) )
def SVDJob(data_path, params):
# test = []
# with open(data_path+'test/test.fold', 'r') as f:
# for line in f:
# test.append( line.strip() )
val_folds = os.listdir(data_path + 'val/')
maes, rmses = [], []
"""HARDCODED AS F**K: 4+1, N20""" #Rationale: N20 por ser folds más chicos. Así, ninguno de estos folds se interlapa con alguno de testing.
for i in range(1, 4 + 1):
svd = pyreclab.SVD(dataset=data_path + 'train/train_N20.' + str(i),
dlmchar=b',',
header=False,
usercol=0,
itemcol=1,
ratingcol=2)
svd.train(factors=params['f'],
maxiter=params['mi'],
lr=params['lr'],
lamb=params['lamb'])
predlist, mae, rmse = svd.test(input_file=data_path + 'val/val_N20.' +
str(i),
dlmchar=b',',
header=False,
usercol=0,
itemcol=1,
ratingcol=2)
maes.append(mae)
rmses.append(rmse)
return mean(maes), mean(rmses)
def test_funksvd():
svd = pyreclab.SVD(dataset='dataset/u1.base',
dlmchar=b'\t',
header=False,
usercol=0,
itemcol=1,
ratingcol=2)
svd.train(factors=1000, maxiter=100, lr=0.01, lamb=0.1)
pred = svd.predict('457', '443')
assert abs(pred - expected_prediction) < prediction_epsilon
ranking = svd.recommend('457', 5, includeRated=False)
assert ranking == expected_ranking
predlist, mae, rmse = svd.test(input_file='dataset/u1.test',
dlmchar=b'\t',
header=False,
usercol=0,
itemcol=1,
ratingcol=2,
output_file='predictions.csv')
assert abs(mae - expected_mae) < mae_epsilon
assert abs(rmse - expected_rmse) < rmse_epsilon
def main(factors=100):
svd = pyreclab.SVD( dataset = 'dataset/u1.base',
dlmchar = b'\t',
header = False,
usercol = 0,
itemcol = 1,
ratingcol = 2 )
print( '-> training model' )
start = time.clock()
svd.train(factors = factors, maxiter = 100, lr = 0.01, lamb = 0.1 )
end = time.clock()
print( 'training time: ' + str( end - start ) )
print( '-> individual test' )
pred = svd.predict( '457', '443' )
print( 'user 457, item 443, prediction ' + str( pred ) )
ranking = svd.recommend( '457', 5, includeRated = False )
print( 'recommendation for user 457: ' + str( ranking ) )
print( '-> prediction test' )
start = time.clock()
predlist, mae, rmse = svd.test( input_file = 'dataset/u1.test',
dlmchar = b'\t',
header = False,
usercol = 0,
itemcol = 1,
ratingcol = 2,
output_file = 'predictions.csv' )
end = time.clock()
print( 'prediction time: ' + str( end - start ) )
print( 'MAE: ' + str( mae ) )
print( 'RMSE: ' + str( rmse ) )
print( '-> recommendation test' )
start = time.clock()
recommendList = svd.testrec( input_file = 'dataset/u1.test',
dlmchar = b'\t',
header = False,
usercol = 0,
itemcol = 1,
ratingcol = 2,
topn = 10,
output_file = 'ranking.json',
includeRated = False )
end = time.clock()
print( 'recommendation time: ' + str( end - start ) )
del svd
time.sleep( 10 )
return mae, rmse
def generate_recommends(params):
svd = pyreclab.SVD(dataset='TwitterRatings/funkSVD/ratings.train',
dlmchar=b',',
header=False,
usercol=0,
itemcol=1,
ratingcol=2)
logging.info("-> Entrenando modelo..")
logging.info(
"N° Factores: {0}; maxiter: {1}; learning rate: {2}; lambda: {3} ".
format(f, mi, lr, lamb))
start = time.clock()
svd.train(factors=params['f'],
maxiter=params['mi'],
lr=params['lr'],
lamb=params['lamb'])
end = time.clock()
logging.info("training time: " + str(end - start))
logging.info("-> Test de Recomendación..")
start = time.clock()
recommendationList = svd.testrec(
input_file='TwitterRatings/funkSVD/ratings.test',
dlmchar=b',',
header=False,
usercol=0,
itemcol=1,
ratingcol=2,
topn=10,
output_file='TwitterRatings/funkSVD/ranking.json')
end = time.clock()
logging.info('recommendation time: ' + str(end - start))
prediction_filename = data_url + 'predictionsSVD ' #5.csv'
ordenada_filename = data_url + 'ordenadasSVD' #5.csv'
print 'SVD'
for i in [1, 5]: #range(1, data_chunks + 1):
f_t = training_filename + str(i) + ".txt"
f_p = probe_filename + str(i) + ".txt"
f_pred = prediction_filename + str(i) + ".csv"
f_ord = ordenada_filename + str(i) + ".csv"
print "Corriendo experimento ", i, "..."
print 'Entrenando...'
obj = pyreclab.SVD(dataset=f_t,
dlmchar=b'\t',
header=True,
usercol=0,
itemcol=1,
ratingcol=2)
obj.train(factors=1000, maxiter=100, lr=0.01, lamb=0.1)
print 'Prediciendo...'
#prediction = obj.predict( "630685", "1")
#ranking = obj.recommend( "630685", 10, True)
#print prediction
#print ranking
predictionList, mae, rmse = obj.test(input_file=f_p,
dlmchar=b'\t',
header=False,
usercol=0,
itemcol=1,
ratingcol=2,
#! /usr/bin/env python
import time
import pyreclab
if __name__ == '__main__':
svd = pyreclab.SVD(dataset='dataset/u1.base',
dlmchar=b'\t',
header=False,
usercol=0,
itemcol=1,
ratingcol=2)
print('-> training model')
start = time.clock()
svd.train(factors=1000, maxiter=100, lr=0.01, lamb=0.1)
end = time.clock()
print('training time: ' + str(end - start))
print('-> individual test')
pred = svd.predict('457', '443')
print('user 457, item 443, prediction ' + str(pred))
ranking = svd.recommend('457', 5)
print('recommendation for user 457: ' + str(ranking))
print('-> prediction test')
start = time.clock()
predlist, mae, rmse = svd.test(input_file='dataset/u1.test',
dlmchar=b'\t',
def svd_tuning(data_path):
defaults = {'f': 1000, 'mi': 100, 'lr': 0.01, 'lamb': 0.1}
results = {'f': {}, 'mi': {}, 'lr': {}, 'lamb': {}}
for param in ['f', 'lamb', 'lr', 'mi']:
if param == 'f':
for i in range(100, 1525, 25):
defaults['f'] = i
logging.info(
"Entrenando con f={f}, lamb={lamb}, lr={lr}, mi={mi}".
format(f=defaults['f'],
lamb=defaults['lamb'],
lr=defaults['lr'],
mi=defaults['mi']))
mae, rmse = SVDJob(data_path=data_path, params=defaults)
results['f'][i] = rmse
defaults['f'] = opt_value(results=results['f'], metric='rmse')
elif param == 'mi':
for i in range(10, 520, 20):
defaults['mi'] = i
logging.info(
"Entrenando con f={f}, lamb={lamb}, lr={lr}, mi={mi}".
format(f=defaults['f'],
lamb=defaults['lamb'],
lr=defaults['lr'],
mi=defaults['mi']))
mae, rmse = SVDJob(data_path=data_path, params=defaults)
results['mi'][i] = rmse
defaults['mi'] = opt_value(results=results['mi'], metric='rmse')
elif param == 'lamb':
for i in [
0.001, 0.005, 0.01, 0.05, 0.1, 0.15, 0.2, 0.25, 0.3, 0.35,
0.4, 0.45, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0
]:
defaults['lamb'] = i
logging.info(
"Entrenando con f={f}, lamb={lamb}, lr={lr}, mi={mi}".
format(f=defaults['f'],
lamb=defaults['lamb'],
lr=defaults['lr'],
mi=defaults['mi']))
mae, rmse = SVDJob(data_path=data_path, params=defaults)
results['lamb'][i] = rmse
defaults['lamb'] = opt_value(results=results['lamb'],
metric='rmse')
elif param == 'lr':
for i in [
0.001, 0.002, 0.003, 0.004, 0.005, 0.006, 0.007, 0.008,
0.009, 0.01, 0.02, 0.03, 0.04, 0.05, 0.06, 0.07, 0.08,
0.09, 0.1
]:
defaults['lr'] = i
logging.info(
"Entrenando con f={f}, lamb={lamb}, lr={lr}, mi={mi}".
format(f=defaults['f'],
lamb=defaults['lamb'],
lr=defaults['lr'],
mi=defaults['mi']))
mae, rmse = SVDJob(data_path=data_path, params=defaults)
results['lr'][i] = rmse
defaults['lr'] = opt_value(results=results['lr'], metric='rmse')
# Real testing
svd = pyreclab.SVD(dataset=data_path + 'eval_train_N20.data',
dlmchar=b',',
header=False,
usercol=0,
itemcol=1,
ratingcol=2)
svd.train(factors=defaults['f'],
maxiter=defaults['mi'],
lr=defaults['lr'],
lamb=defaults['lamb'])
predlist, mae, rmse = svd.test(input_file=data_path + 'test/test_N20.data',
dlmchar=b',',
header=False,
usercol=0,
itemcol=1,
ratingcol=2)
with open('TwitterRatings/funkSVD/opt_params_ptcdssplit.txt', 'w') as f:
for param in defaults:
f.write("{param}:{value}\n".format(param=param,
value=defaults[param]))
f.write("RMSE:{rmse}, MAE:{mae}".format(rmse=rmse, mae=mae))
return defaults
def PRF_calculator(params, folds, topN):
ratings_train, ratings_test = [], []
with open('TwitterRatings/funkSVD/ratings.train', 'r') as f:
for line in f:
ratings_train.append(line.strip())
with open('TwitterRatings/funkSVD/ratings.test', 'r') as f:
for line in f:
ratings_test.append(line.strip())
preferred_consumption = consumption(
ratings_path='TwitterRatings/funkSVD/ratings.test',
rel_thresh=4,
with_ratings=False)
for n in topN:
precision_folds, recall_folds = [], []
# for _ in range(0, folds):
# ratingsSampler(ratings_train, 'TwitterRatings/funkSVD/ratings_temp.train', 0.8)
# ratingsSampler(ratings_test, 'TwitterRatings/funkSVD/ratings_temp.test', 0.8)
svd = pyreclab.SVD(
dataset=
'TwitterRatings/funkSVD/ratings.train', #o ratings_temp.train
dlmchar=b',',
header=False,
usercol=0,
itemcol=1,
ratingcol=2)
svd.train(factors=params['f'],
maxiter=params['mi'],
lr=params['lr'],
lamb=params['lamb'])
recommendationList = svd.testrec(
input_file=
'TwitterRatings/funkSVD/ratings.test', #o ratings_temp.test
dlmchar=b',',
header=False,
usercol=0,
itemcol=1,
ratingcol=2,
topn=n,
includeRated=False)
users_precisions, users_recalls = [], []
for userId in recommendationList[0]:
recs = set(recommendationList[0][userId])
cons = set(preferred_consumption[userId])
tp = len(recs & cons)
fp = len(recs - cons)
fn = len(cons - recs)
users_precisions.append(float(tp) / (tp + fp))
try:
users_recalls.append(float(tp) / (tp + fn))
except ZeroDivisionError as e:
continue
precision_folds.append(mean(users_precisions))
recall_folds.append(mean(users_recalls))
p = mean(precision_folds)
r = mean(recall_folds)
f = 2 * p * r / (p + r)
with open('TwitterRatings/funkSVD/recall.txt', 'a') as file:
file.write("N=%s, P=%s, R=%s, F=%s\n" % (n, p, r, f))
import time
import pyreclab
if __name__ == '__main__':
model = pyreclab.SVD(factors=1000,
dataset='dataset/u1.base',
dlmchar=b'\t',
header=False,
usercol=0,
itemcol=1,
ratingcol=2)
print('-> training model')
start = time.clock()
model.train(maxiter=100, lr=0.01, lamb=0.1, progress=True)
end = time.clock()
print('training time: ' + str(end - start))
print('-> individual test')
pred = model.predict('457', '443')
print('user 457, item 443, prediction ' + str(pred))
ranking = model.recommend('457', 5, includeRated=False)
print('recommendation for user 457: ' + str(ranking))
print('-> prediction test')
start = time.clock()
predlist, mae, rmse = model.test(input_file='dataset/u1.test',
dlmchar=b'\t',
header=False,