def plot_cnot_prob(cnot):
probs = cnot[0]
x = [1, 2, 3, 4]
rects = plt.bar(x, probs)
for rect in rects:
height = rect.get_height()
plt.text(rect.get_x() + rect.get_width() / 2.,
1.05 * height,
f'{height:.3f}',
ha='center',
va='bottom')
plt.xticklabels(['|00>', '|01>', '|10>', '|11>'])
plt.set_ylim(0., 1.4)
plt.show()
def histogram(samples, xlabels, name=''):
# distribution of examples per class displayed as a histogram
plt.rcParams['figure.figsize'] = (8, 6)
n, bins, patches = plt.hist(samples,
len(xlabels),
facecolor='green',
histtype='bar',
rwidth=0.8,
alpha=1)
plt.grid(True)
plt.xticklabels(xlabels)
if name and not name.isspace():
plt.savefig(name)
def main():
fig = plt.figure(figsize=(15, 10))
data = {}
files = []
for file in glob.glob("*.dat"):
files.append(file)
files = sorted(files, key=lambda x: int(x.split('_')[-1].split('.')[0]))
for file in files:
with open(file, 'r') as f:
file_number = int(file.split('_')[-1].split('.')[0])
data[file_number] = []
for line in f.readlines():
dat = line.split()[-1]
data[file_number].append(float(dat))
# plot and annotate histograms
y, x, _ = plt.hist(data[file_number], bins=100, histtype='step')
y_coord = max(y)
x_coord = x[np.where(y == max(y))[0][0]]
plt.annotate(str(file_number), (x_coord, y_coord), ha='center')
plt.xlabel('dihedral angle')
plt.ylabel('frequency')
xticks = np.arange(-130, 50, 15)
plt.xticks(xticks)
plt.xticklabels([str(k) + "°" for k in xticks])
# use quantiles to check proper overlap of distributions
with open('umbrella_stats.txt', 'w') as o:
o.write('file number <-> filenumber : sufficient overlap\n')
for key in sorted(data.keys()):
key_next = key + 1
if key_next in data.keys():
dat = np.asarray(data[key])
dat_next = np.asarray(data[key_next])
dat_quantile = np.quantile(dat, 0.8)
dat_next_quantile = np.quantile(dat_next, 0.2)
# plt.axvline(dat_quantile, color='red')
# plt.axvline(dat_next_quantile)
if (dat_quantile > dat_next_quantile):
o.write('{} <-> {} : True\n'.format(key, key_next))
else:
print('!!! {} <-> {} : False\n'.format(key, key_next))
o.write('{} <-> {} : False\n'.format(key, key_next))
plt.savefig('histograms.png')
def sexyaxis(*args, **kwargs):
a = xlim
minf = a[1]
maxf = a[2]
possibles = arange(1, 50) * 2000
if (maxf - minf) < 8000:
possibles = possibles / 2
ticks = array([])
labels = []
for i in range(1, size(possibles)):
if (possibles[i] >= minf) and (possibles[i] <= maxf):
ticks[end + 1] = possibles[i]
labels[end + 1] = print('%d' % possibles[i])
mpl.xticks(ticks)
mpl.xticklabels(labels)
return a
metrics_ = []
for k in range(1000):
X2_training, X2_testing, y2_training, y2_testing = train_test_split(X2, y_true, test_size=test_size)
clf = GradientBoostingClassifier(**params)
clf.fit(X2_training[:,89:], y2_training)
metrics_ += [compute_metrics(y2_testing, clf.predict(X2_testing[:,89:]))]
print_metrics(y2_testing, clf.predict(X2_testing[:,89:]))
plt.xlabel('iteration')
plt.ylabel('Training loss')
plt.title('Training loss by considering just one frame 18h before eruption')
plt.plot(clf.train_score_)
plt.show()
plt.xlabel('Features')
plt.xticklabels()
plt.ylabel('Relative importance')
plt.plot(clf.feature_importances_, '+')
plt.show()
plt.xlabel('Nb of frames considered (time points)')
plt.ylabel('Accuracy')
# SVM (rbf kernel)
scaler = StandardScaler()
scaler.fit(X2_training)
# Best performance (acc, tss) for X_train: gamma = 7, kernel = rbf, C = 50
# Best performance (tss) for X2_train: gamma = 0.037, C = 24
clf = svm.SVC(C = 24, kernel='rbf', gamma = 0.037, tol = 1e-10)
clf.fit(scaler.transform(X2_training), y2_training)
scaler.fit(X2_testing)
.max(),
where=(df['data'] > '01/01') & (df['data'] < '03/31'),
facecolor='blue',
alpha=0.5)
plt.fill_between(
df['data'],
0,
df['PSZ-AC:1 HEAT PUMP DX COOLING COIL:Cooling Coil Total Cooling Rate [W](TimeStep)']
.max(),
where=(df['data'] > '11/02') & (df['data'] < '12/31'),
facecolor='blue',
alpha=0.5)
plt.legend()
plt.xlabel('Date')
plt.xticks(df['data'][::30])
plt.xticklabels(df['data'][::30], rotation=45)
# plt.xticks(df['data'])
plt.title('Heat pump Heating and Cooling coil rate')
plt.show()
# Prova 2: con ax.
fig, ax = plt.subplots()
ax.bar(
df['data'],
df['PSZ-AC:1 HEAT PUMP DX HEATING COIL:Heating Coil Heating Rate [W](TimeStep)'],
color='blue',
label='Coil Heating Rate [W]')
# df.plot(x='data', y=['PSZ-AC:1 HEAT PUMP DX COOLING COIL:Cooling Coil Total Cooling Rate [W](TimeStep)', 'PSZ-AC:1 HEAT PUMP DX HEATING COIL:Heating Coil Heating Rate [W](TimeStep)'])
ax.bar(
df['data'],
df['PSZ-AC:1 HEAT PUMP DX COOLING COIL:Cooling Coil Total Cooling Rate [W](TimeStep)'],
ax[1].set_xlabel('Time (ms)')
ax[1].set_ylabel('Channel')
ax[1].axvline(x=30, color='black')
ax[1].set_title('larger boundary ' + band)
fig.subplots_adjust(right=0.8)
cbar_ax = fig.add_axes([0.85, 0.15, 0.03, 0.7])
fig.colorbar(img, cax=cbar_ax)
plotbydrift('highbeta', 2)
plt.legend()
x_label_list = np.arange(-300, 600, 100)
plt.xticks(np.arange(0, 90, 10))
plt.xticklabels(x_label_list)
plt.set_xlabel('Time (ms)')
plt.set_ylabel('Power')
def runplot(var1):
easy, med, hard = np.arange(0, 33), np.arange(33, 66), np.arange(66, 99)
condlist = [easy, med, hard]
titlelist = ['easy', 'med', 'hard']
fig, ax = plt.subplots(1, 3, figsize=(17, 5))
fig.tight_layout(pad=2)
for i in range(0, 3):
cond = condlist[i]
# limit = max(abs(np.min(theta_fast)), abs(np.max(theta_fast)))
mat = np.mean(band_dic[var1][cond], axis=0)
def make_plot(data_to_plot,
colors_to_plot,
plot_title,
filename,
sub_dir='',
xlim=(-50, 50),
xticks=[-40, -20, 0, 20, 40],
ylim=(29, 71),
yticks=[30, 40, 50, 60, 70],
yline=50,
plot_yline=True,
legend_labels=[],
xtick_labels=[],
xlabel='Distance to TSS (kb), Txn L-->R',
ylabel='% of forks moving L-->R',
legend_loc='lower right'):
plt.figure(figsize=(fig_size_width, fig_size_height), dpi=300)
for i, data in enumerate(data_to_plot):
color = colors_to_plot[i]
if (len(legend_labels) > 0):
lab = legend_labels[i]
else:
lab = ''
plt.plot(
range(int(-1 * length_around_site / 2),
int(length_around_site / 2 + 1)),
data,
'-',
color=color,
label=lab,
linewidth=1.25,
)
if (len(ylim) == 2):
plt.ylim(ylim)
if (len(yticks) > 0):
plt.yticks(yticks)
if (len(xlim) == 2):
plt.xlim(xlim)
if (len(xticks) > 0):
plt.xticks(xticks)
if (len(xtick_labels) > 0):
plt.xticklabels(xtick_labels)
plt.title(plot_title, x=0.02, y=0.9, ha='left', va='top')
if (plot_yline == True):
plt.axhline(y=yline, color='black', linewidth=0.8, dashes=[2, 4])
plt.axvline(linestyle='--', color='black', linewidth=0.8, dashes=[2, 4])
if (len(legend_labels) > 0):
if (legend_loc == 'lower right'):
legend = plt.legend(loc='lower right',
fancybox=False,
bbox_to_anchor=(1.03, -0.035))
elif (legend_loc == 'upper right'):
legend = plt.legend(loc='upper right',
fancybox=False,
bbox_to_anchor=(1.03, 1.035))
else:
legend = plt.legend(loc=legend_loc)
frame = legend.get_frame()
frame.set_linewidth(0.6)
frame.set_edgecolor('black')
plt.xlabel(xlabel)
plt.ylabel(ylabel)
plt.tight_layout(pad=2)
plt.savefig(base_dir + '/' + sub_dir + '/' + filename + '.' + ext,
transparent=True)
plt.clf()
plt.close()