def run_benchmark(self, output_file_path):
"""
Run benchmark for all learners on every dataset provided.
The result will be output to the provided file.
"""
dataloader = DataLoader()
f = open(output_file_path,'w')
# score, learner_1_name, learner_2_name, ..., learner_n_name
f.write("dataset")
for learner in self.learners:
f.write(", " + learner.name())
# write scores for all data sets
for dataset_id in self.dataset_ids:
print "Benchmarking dataset: " + str(dataset_id)
f.write("\n" + str(dataset_id))
train_data = dataloader.load_sequences_from_file("../data/" + str(dataset_id) + ".pautomac" + ".train")
test_data = dataloader.load_sequences_from_file("../data/" + str(dataset_id) + ".pautomac" + ".test")
solution_data = dataloader.load_probabilities_from_file("../data/" + str(dataset_id) + ".pautomac_solution" + ".txt")
for learner in self.learners:
print "Training learner: " + learner.name()
learner.train(train_data, test_data)
print "Evaluating learner: " + learner.name()
score = learner.evaluate(test_data, solution_data)
print "Achieved score: " + str(score)
str_score = " {0:.1f}".format(score)
while len(str_score) < 8:
str_score = " " + str_score
f.write(", " + str_score)
f.close()
from numpy import *
from decimal import *
from sys import *
from learner import Learner
from decimal import *
from sys import *
from utilities import *
from dataLoader import DataLoader
import time
list1 = [[1, 2], [3, 4], [5, 6]]
list2 = [2, 3]
#for x in xrange(0, len(list1), 2):
# print list1[x]
dataloader = DataLoader()
train_data = dataloader.load_sequences_from_file("../data/" + "1" + ".pautomac" + ".test")
#comps = collect_unique_symbol_compositions(train_data, 2)
MathiasLearner.train(train_data)
#print comps.index([1, 1])
