def means(X):
mean = []
for col in range(len(X[0])):
sum_val = 0.0
for rows in range(len(X)):
sum_val = sum_val + X[rows][col]
mean.append(sum_val / len(X))
return mean
def result(datapath, algorithm):
result_value = []
for subjects in os.listdir(datapath):
subject_path = os.path.join(datapath, subjects)
for tasks in os.listdir(subject_path):
mean = []
maxi = []
mini = []
var = []
ent = []
subject_list = []
task_list = []
subject_list.append(subjects)
task_path = os.path.join(subject_path, tasks)
if tasks == 'T8':
tasks = '1'
else:
tasks = '0'
task_list.append(tasks)
for files in os.listdir(task_path):
if files != '.DS_Store' and files != 'ECG_mV.txt':
file_path = os.path.join(task_path, files)
data = np.loadtxt(file_path, dtype='float')
Mean = np.mean(data)
mean.append(Mean)
Variance = np.var(data)
var.append(Variance)
Minimum = min(data)
mini.append(Minimum)
Maximum = max(data)
maxi.append(Maximum)
Entropy = max(data)
ent.append(Entropy)
output = mean + var + mini + maxi + ent
result_name = subject_list + task_list + output
result_value.append(result_name)
with open('results.csv', 'w') as myfile:
wr = csv.writer(myfile, lineterminator='\n')
wr.writerows(result_value)
if algorithm == 'RF':
randomForest()
else:
svm()
def normalize_data(data):
std = []
mean = []
dataOut = data.copy
for i in range(dataOut.shape[0]):
std.append([np.std(dataOut, ddof=1) for j in range(dataOut.shape[0])])
mean.append([np.mean(dataOut) for k in range(dataOut.shape[0])])
std = list(itertools.chain(*std))
mean = list(itertools.chains(*mean))
for i in range(data.shape[0]):
dataOut[i, :] = (dataOut[i, :] - mean[i]) / std[i]
dataOut[np.isnan(dataOut)] = 0
return dataOut
def word_averaging(wv, words):
all_words, mean = set(), []
for word in words:
if isinstance(word, np.ndarray):
mean.append(word)
elif (word == 'NoneType'):
print(word)
elif word in wv.vocab:
mean.append(wv.syn0norm[wv.vocab[word].index])
all_words.add(wv.vocab[word].index)
if not mean:
return np.zeros(wv.vector_size, )
mean = gensim.matutils.unitvec(np.array(mean).mean(axis=0)).astype(
np.float32)
return mean
def box_graph(means, times, data):
mean = []
time = []
for i in range(4):
mean.append((means[i][0], means[i][1], means[i][2], means[i][3]))
time.append((times[i][0], times[i][1], times[i][2], times[i][3]))
for i in range(4):
sns.set()
sns.set_style("whitegrid", {'grid.linestyle': '--'})
sns.set_context("paper", 1.5, {"lines.linewidth": 4})
sns.set_palette("winter_r", 8)
sns.set('talk',
'whitegrid',
'dark',
rc={
"lines.linewidth": 2,
'grid.linestyle': '--'
})
fig, ax = plt.subplots()
bp = ax.boxplot(mean[i], vert=True, patch_artist=True)
# bp = ax.boxplot(time[i], vert=True, patch_artist=True)
for box in bp['boxes']:
box.set(color="black", linewidth=1.5)
for box in bp['medians']:
plt.setp(box, color="black", linewidth=1.5)
for box in bp['caps']:
plt.setp(box, color="black", linewidth=1.5)
for box in bp['whiskers']:
plt.setp(box, ls="solid", color="black", linewidth=1.5)
for box, color in zip(bp["boxes"], sns.color_palette("Set3", 6)):
box.set_facecolor(color)
ax.set_xticklabels(['ST-GIB', 'CD-GIB', 'FD-GIB', 'TR-GIB'])
plt.xlabel('layout')
plt.ylabel('accuracy [%]')
# plt.ylabel('completion time [ms]')
# plt.ylim(1000, 10000)
plt.ylim(0, 110)
# ax.legend(bp["boxes"], ['ST-GIB', 'CD-GIB', 'FD-GIB', 'TR-GIB'], loc='upper right')
plt.legend()
plt.grid()
plt.savefig('../src/trajectory/mean' + str(i + 1) + '.png')
# plt.savefig('../src/trajectory/time' + str(i+1) + '.png')
plt.close()
def evolve(self,generations = 550):
#generations = 550
gen = []
best = []
mean = []
found = False
cur_best,cur_mean = 0,0
for i in range(generations):
self.sortIndivids()
cur_best,cur_mean = self.getBestAndMean()
gen.append(i)
best.append(cur_best)
mean.append(cur_mean)
if cur_best==0 and not found:
print('found solution in generation {}!\n'.format(i))
self.sorted_population[0][0].printState()
found = True
self.mateGrid()
date_string = datetime.now().strftime("%H-%M-%S")
print('\n\nending pop:\n')
[print(tuple[1],tuple[0].state) for tuple in self.sorted_population]
print('\nending mean:',cur_mean)