def run(xls_file, value_file, itemset_str_lst, delimiter, alternative):
global readFile
import readFile
transaction_list, columnid2name = readFile.readFiles(
xls_file, value_file, delimiter)
if alternative < 0:
global lamp
from lamp import reverseValue
transaction_list = reverseValue(transaction_list, "u_test")
func = FunctionOfX(transaction_list, alternative)
colname2id_dict = readFile.colname2id(columnid2name)
itemset = set()
for i in itemset_str_lst:
item_id = colname2id_dict[i]
itemset.add(item_id + 1)
flag_transactions_id = []
for i in range(len(transaction_list)):
t = transaction_list[i]
if len(itemset & t.itemset) == len(itemset):
flag_transactions_id.append(i)
p_value, stat_score = func.calPValue(transaction_list,
flag_transactions_id)
# print i
# print item_id
# print columnid2name[item_id]
# p, stat_score = func.calPValue(transaction_list, itemset)
n = len(transaction_list)
sys.stdout.write("p-value: %g (N: %s, x: %s, z-score: %f)\n" \
% (p_value, n, len(flag_transactions_id), stat_score))
return (p_value, len(flag_transactions_id))
def run(xls_file, value_file, itemset_str_lst, delimiter, alternative):
global readFile
import readFile
transaction_list, columnid2name = readFile.readFiles(
xls_file, value_file, delimiter)
max_lambda = maxLambda(transaction_list)
if alternative < 0:
global lamp
from lamp import reverseValue
transaction_list = reverseValue(transaction_list, "fisher")
func = FunctionOfX(transaction_list, max_lambda, abs(alternative))
colname2id_dict = readFile.colname2id(columnid2name)
itemset = set()
for i in itemset_str_lst:
item_id = colname2id_dict[i]
itemset.add(item_id + 1)
flag_transactions_id = []
for i in range(len(transaction_list)):
t = transaction_list[i]
if len(itemset & t.itemset) == len(itemset):
flag_transactions_id.append(i)
p_value, stat_value = func.calPValue(transaction_list,
flag_transactions_id)
n = len(transaction_list)
n1 = func.getN1()
sys.stdout.write("p-value: %s (N: %s, n1: %s, x: %s, a: %s)\n" %
(p_value, n, n1, len(flag_transactions_id), stat_value))
return (p_value, len(flag_transactions_id))
def run(xls_file, value_file, itemset_str_lst, delimiter, alternative):
global readFile
import readFile
transaction_list, columnid2name = readFile.readFiles(xls_file, value_file, delimiter)
max_lambda = maxLambda(transaction_list)
if alternative < 0:
global lamp
from lamp import reverseValue
transaction_list = reverseValue( transaction_list, "fisher" )
func = FunctionOfX(transaction_list, max_lambda, abs( alternative ) )
colname2id_dict = readFile.colname2id(columnid2name)
itemset = set()
for i in itemset_str_lst:
item_id = colname2id_dict[i]
itemset.add(item_id + 1)
flag_transactions_id = []
for i in xrange( len(transaction_list) ):
t = transaction_list[i]
if len( itemset & t.itemset ) == len(itemset):
flag_transactions_id.append( i )
p_value, stat_value = func.calPValue(transaction_list, flag_transactions_id)
n = len(transaction_list)
n1 = func.getN1()
sys.stdout.write("p-value: %s (N: %s, n1: %s, x: %s, a: %s)\n"
% (p_value, n, n1, len(flag_transactions_id), stat_value))
return (p_value, len(flag_transactions_id))
def run(xls_file, value_file, itemset_str_lst, delimiter, alternative):
global readFile
import readFile
transaction_list, columnid2name = readFile.readFiles(xls_file, value_file, delimiter)
if alternative < 0:
global lamp
from lamp import reverseValue
transaction_list = reverseValue(transaction_list, "u_test")
func = FunctionOfX(transaction_list, alternative)
colname2id_dict = readFile.colname2id(columnid2name)
itemset = set()
for i in itemset_str_lst:
item_id = colname2id_dict[i]
itemset.add(item_id + 1)
flag_transactions_id = []
for i in xrange(len(transaction_list)):
t = transaction_list[i]
if len(itemset & t.itemset) == len(itemset):
flag_transactions_id.append(i)
p_value, stat_score = func.calPValue(transaction_list, flag_transactions_id)
# print i
# print item_id
# print columnid2name[item_id]
# p, stat_score = func.calPValue(transaction_list, itemset)
n = len(transaction_list)
sys.stdout.write("p-value: %g (N: %s, x: %s, z-score: %f)\n" % (p_value, n, len(flag_transactions_id), stat_score))
return (p_value, len(flag_transactions_id))
def run(transaction_file, flag_file, threshold, k, set_method, lcm_path, max_comb, log_file, alternative):
# read 2 files and get transaction list
sys.stderr.write( "Read input files ...\n" )
transaction_list = set()
try:
transaction_list, columnid2name = readFile.readFiles(transaction_file, flag_file, ",")
# If the alternative hypothesis is 'less',
# the positive and negative of observe values are reversed,
# and conduct the identical procedure to 'greater'.
if alternative < 0:
transaction_list = lamp.reverseValue( transaction_list, set_method )
max_comb = lamp.convertMaxComb( max_comb, len(columnid2name) )
except ValueError as e:
return
except KeyError as e:
return
trans4lcm = transaction_file + ".4lcm53" # the filename for outputting logs
# run multiple test
try:
outlog = open( log_file, 'w' )
except IOError as e:
outlog.close()
start_time = time.time()
# generate null distribution
sys.stderr.write( "Calculate the minimum p-value distribution using the permutation test ...\n" )
outlog.write("Calculate the minimum p-value distribution using the permutation test ...\n")
min_p_list, fre_pattern, func_f = \
generateMinPDist(transaction_list, trans4lcm, threshold, set_method, \
lcm_path, max_comb, k, outlog, alternative)
# adjusted significance level
outlog.write("Adjust significance level ...\n")
adjusted_threshold, sorted_min_p_list = adjustedThreshold( min_p_list, threshold, k )
outlog.write("Adjusted significance level: %s\n" % adjusted_threshold)
correction_term_time = time.time()
# enumerate combination whose P-value up to adjusted threshold
outlog.write("Calculate the p-values in the given data set ...\n")
enrich_lst, time_enumerate_freq, time_enumerate_total = \
enumerateSigComb( transaction_list, trans4lcm, fre_pattern, func_f, \
max_comb, adjusted_threshold, outlog )
finish_test_time = time.time()
# output the significant combinations
outputResult( transaction_file, flag_file, threshold, k, set_method, max_comb, columnid2name, \
enrich_lst, adjusted_threshold, transaction_list, func_f, sorted_min_p_list, alternative )
# output time cost
sys.stdout.write("Time (sec.): Computing correction factor %.3f, Enumerating significant combinations %.3f, Total %.3f\n" \
% (correction_term_time-start_time, time_enumerate_total, finish_test_time - start_time))
# output the minimum P-values
outputMinP( min_p_list )
outlog.close()
return enrich_lst, adjusted_threshold, columnid2name
def run(transaction_file, flag_file, threshold, k, set_method, lcm_path, max_comb, log_file, alternative): # read 2 files and get transaction list sys.stderr.write( "Read input files ...\n" ) transaction_list = set() try: transaction_list, columnid2name = readFile.readFiles(transaction_file, flag_file, ",") # If the alternative hypothesis is 'less', # the positive and negative of observe values are reversed, # and conduct the identical procedure to 'greater'. if alternative < 0: transaction_list = lamp.reverseValue( transaction_list, set_method ) max_comb = lamp.convertMaxComb( max_comb, len(columnid2name) ) except ValueError, e: return
def run(transaction_file, flag_file, threshold, k, set_method, lcm_path,
max_comb, log_file, alternative):
# read 2 files and get transaction list
sys.stderr.write("Read input files ...\n")
transaction_list = set()
try:
transaction_list, columnid2name = readFile.readFiles(
transaction_file, flag_file, ",")
# If the alternative hypothesis is 'less',
# the positive and negative of observe values are reversed,
# and conduct the identical procedure to 'greater'.
if alternative < 0:
transaction_list = lamp.reverseValue(transaction_list, set_method)
max_comb = lamp.convertMaxComb(max_comb, len(columnid2name))
except ValueError, e:
return
