def plotting(xgrid, vgrid, x, v, f, scalar_field, f0, timestep, plot_save,
plot_show):
if (plot_show == 1):
plt.clf()
plt.ion()
# plt.subplot(2,1,1)
plt.pcolormesh(xgrid, vgrid, f)
plt.axis([x[0], x[-1], v[0], v[-1]])
plt.xlabel('X', fontsize=20)
plt.ylabel('V', fontsize=20)
plt.title('Phase Space')
# Electric Field Comparison
# time = 0.02*timestep
# Et = 4.0*0.01 * 0.3677 * np.exp(-0.1533*time)*np.sin(0.5*x)*np.cos(1.4156*time - 0.5326245)
# plot_title = 't = '+str(timestep*0.02)
# plt.plot(x,scalar_field)
# plt.plot(x,Et)
# plt.title(plot_title)
# plt.subplot(2,1,2)
# plt.plot(x,potential)
# plt.xlabel('X')
# plt.ylabel('$\phi$')
if (plot_show == 1):
plt.draw()
plt.pause(.05)
if (plot_save == 1):
FIG_NAME = 'OUT/f_nonlinear_' + str(timestep) + '.png'
plt.savefig(FIG_NAME)
def plot(self):
self._fig.clf()
self._ax.cla()
self._fig = plt.figure('taylor')
self._ax = self._fig.add_subplot(1, 1, 1)
self._ax.plot(self._x_list,
self._y_list,
linestyle='solid',
label='origin')
for order in range(len(self._taylor_y_array)):
label = '{} order approximation'.format(order + 1)
self._ax.plot(self._x_list,
self._taylor_y_array[order],
linestyle='dashed',
label=label)
self._ax.plot([self.base_x],
self._model.fx([self.base_x]),
marker='.',
label='approximation base point')
self._ax.set_title('Taylor expansion')
self._ax.legend(loc='lower right')
self._ax.set_ylim(min(self._y_list), max(self._y_list))
plt.pause(0.01)
def on_epoch_end(self, epoch, logs={}):
self.val_mse_acc.append(logs.get('val_mean_squared_error'))
self.val_mae_acc.append(logs.get('val_mean_absolute_error'))
self.mse_acc.append(logs.get('mean_squared_error'))
self.mae_acc.append(logs.get('mean_absolute_error'))
self.losses.append(logs.get('loss'))
self.val_losses.append(logs.get('val_loss'))
self.x.append(self.i)
self.i += 1
self.ax1.cla()
self.ax2.cla()
self.ax3.cla()
self.ax1.set_title('Losses')
self.ax1.set(ylabel='Logarithmic MSE')
self.ax2.set_title('Accuracy (MSE)')
self.ax2.set(ylabel='Mean Squared Error')
self.ax3.set_title('Accuracy (MAE)')
self.ax3.set(ylabel='Mean Absolute Error')
self.ax1.plot(self.x, self.losses, label="loss")
self.ax1.plot(self.x, self.val_losses, label="val_loss")
self.ax1.legend(loc="upper right")
self.ax2.plot(self.x, self.mse_acc, label="training")
self.ax2.plot(self.x, self.val_mse_acc, label="validation")
self.ax2.legend(loc="upper right")
self.ax3.plot(self.x, self.mae_acc, label="training")
self.ax3.plot(self.x, self.val_mae_acc, label="validation")
self.ax3.legend(loc="upper right")
plt.pause(.01)
plt.draw()
plt.savefig("lossaccplot.png", dpi='figure')
def movieshow(self):
self.var = self.dlg.comboBoxSpatialVariables.currentText()
if self.var == 'Not Specified':
QMessageBox.critical(self.dlg, "Error", "No variable is selected")
return
cmin = self.dlg.spinBoxMovieMin.value()
cmax = self.dlg.spinBoxMovieMax.value()
plt.figure(1, figsize=(9, 9), facecolor='white')
plt.ion()
index = 0
for file in self.l:
if file.startswith(self.var + '_') and file.endswith('.tif'):
self.posAll.append(index)
index += 1
index = 0
for i in self.posAll:
gdal_dsm = gdal.Open(self.folderPath[0] + '/' + self.l[i])
grid = gdal_dsm.ReadAsArray().astype(np.float)
plt.imshow(grid, clim=(cmin, cmax))
plt.title(self.l[i])
if index == 0:
plt.colorbar(orientation='horizontal')
plt.show()
plt.pause(0.5)
index += 1
del self.posAll[:]
def histogramac(histc):
histc=histc.histogram(255)
B.show()
pylab.plot(histc)
pylab.draw()
pylab.pause(0.0001)
def wait(seconds):
"""Wait for specified seconds.
Args:
seconds: Time to wait. (waiting only the half to simulate faster)
"""
plt.pause(seconds / 2)
def draw_graph(theta): #theta = math.pi/2 - theta t = np.linspace(0, 2, num=500) # Set time as 'continous' parameter. #for i in theta: # Calculate trajectory for every angle x1 = [] y1 = [] for k in t: x = ((v*k)*math.cos(theta)) y = ((v*k)*math.sin(theta))-((0.5*9.81)*(k*k)) x1.append(x) y1.append(y) p = [i for i, j in enumerate(y1) if j < 0] for i in sorted(p, reverse = True): del x1[i] del y1[i] plot.axis([0.0,2.5, 0.0,2.0]) ax = plot.gca() ax.set_autoscale_on(False) theta_deg = math.degrees(theta) plot.plot(x1, y1, label='yServoAngle = %f' % theta_deg) # Plot for every angle plot.legend(loc='upper right') plot.ion() plot.pause(0.000000001) plot.draw() # And show on one graphic graph1.clear()
def sch_distPdist_linkage_fcluster(points, t, method='complete', metric='euclidean'):
# 1. 层次聚类
# 生成点与点之间的距离矩阵,这里用的欧氏距离:
disMat = sch.distance.pdist(points, metric)
#print(disMat)
# 进行层次聚类:
Z = sch.linkage(disMat, method, metric)
# # 将层级聚类结果以树状图表示出来并保存为plot_dendrogram.png
# P = sch.dendrogram(Z)
# plt.savefig('plot_dendrogram.png')
# # 根据linkage matrix Z得到聚类结果:
cluster = sch.fcluster(Z, t, criterion='maxclust')
#print("Original cluster by hierarchy clustering:\n", cluster)
# PCA 主成份分析 用于结果的散点图显示
pca = PCA(2) # 选取2个主成份
pca.fit(points)
low_d = pca.transform(points) # 降低维度
plt.figure()
mark = ['pb', 'or', 'ob', 'og', 'ok', 'oy', 'om', 'oc',
'sr', 'sb', 'sg', 'sk', 'sy', 'sm', 'sc',
'pr', 'pb', 'pg', 'pk', 'py', 'pm', 'pc',
'Dr', 'Db', 'Dg', 'Dk', 'Dy', 'Dm', 'Dc']
for i in range(len(points)):
markIndex = int(cluster[i])%28+1 # 为样本指定颜色
plt.plot(low_d[i, 0], low_d[i, 1], mark[markIndex], markersize=6)
#plt.text(points[i, 0], points[i, 1], i)
plt.grid()
plt.show()
plt.pause(3)
return cluster
def draw_epsilon(criteria,
initial_sample_size,
batch_size,
color,
fontsize=24,
isLegend=True):
plt.figure(3, [8, 5])
plt.clf()
plt.ylim(-0.1, 1.1)
for criterion in criteria:
sample_size_list = np.arange(
batch_size + initial_sample_size,
batch_size * (criterion.error_ratio.shape[0] + 1) +
initial_sample_size, batch_size)
stopping_data_size = batch_size + initial_sample_size + criterion.stop_timings * batch_size
plt.plot(sample_size_list, criterion.error_ratio, c="k")
plt.plot([
batch_size + initial_sample_size,
batch_size * criterion.error_ratio.shape[0] + initial_sample_size
], [criterion.threshold, criterion.threshold],
c=color[criterion.criterion_name],
label=r"$\lambda={0}$".format(criterion.threshold))
plt.plot([stopping_data_size, stopping_data_size],
[0, criterion.error_ratio.max()],
c=color[criterion.criterion_name],
linestyle="dashed")
if isLegend:
plt.legend(fontsize=fontsize)
plt.xlabel("Sample size", fontsize=fontsize)
plt.ylabel("Error ratio", fontsize=fontsize)
plt.tick_params(labelsize=fontsize)
plt.pause(0.01)
def draw_gene_error(test_error,
criteria,
initial_sample_size,
batch_size,
color,
fontsize=24,
isLegend=True):
plt.figure(1, [8, 5])
plt.clf()
sample_size_list = np.arange(
batch_size + initial_sample_size,
batch_size * (test_error.shape[0] + 1) + initial_sample_size,
batch_size)
plt.plot(sample_size_list, test_error, c='k', label="Generalization error")
for criterion in criteria:
stopping_data_size = batch_size + initial_sample_size + criterion.stop_timings * batch_size
plt.plot([stopping_data_size, stopping_data_size],
[test_error.min(), test_error.max()],
c=color[criterion.criterion_name],
linestyle="dashed",
label=r"$\lambda={0}$".format(criterion.threshold))
if isLegend:
plt.legend(fontsize=fontsize)
plt.xlabel("Sample size", fontsize=fontsize)
plt.ylabel("Generalization error", fontsize=fontsize)
plt.tick_params(labelsize=fontsize)
plt.pause(0.01)
def draw_correlations(result,
y_label,
data_names,
color,
label,
fontsize=24,
ylim=1):
fig = plt.figure(4, [25, 10])
plt.clf()
plt.ylim(0, ylim)
y_ticks = np.linspace(0, ylim, 11)
sum_left = 0
lefts = []
for i, data_name_list in enumerate(data_names):
left = np.arange(len(data_name_list)) + sum_left
plt.bar(left,
result[i],
color=color[i],
label=label[i],
align="center",
ecolor="k")
sum_left += len(data_name_list)
lefts.append(left)
left = np.concatenate(lefts)
data_names = list(itertools.chain.from_iterable(data_names))
plt.xticks(left, data_names)
fig.autofmt_xdate(rotation=20)
plt.yticks(y_ticks)
plt.ylabel(y_label, fontsize=fontsize)
plt.tick_params(labelsize=fontsize)
plt.legend(fontsize=fontsize, loc='lower left')
plt.tight_layout()
plt.grid(axis='y', linestyle='dotted', color='k')
plt.pause(0.01)
def plot(self, cmap='Greys', pause=1):
if self.fig is None:
self.fig = plt.figure(1, figsize=(24, 5))
self.ax1 = self.fig.add_subplot(211)
self.ax2 = self.fig.add_subplot(212)
plt.tight_layout()
#plt.show()
else:
plt.figure(1)
#... build numpy array... FOR EXAMPLE :
aview = np.random.randint(0, 2, (5, 300))
pview = np.random.randint(0, 2, (5, 300))
self.ax1.imshow(aview,
cmap=cmap,
interpolation='nearest',
aspect='auto')
self.ax2.imshow(pview,
cmap=cmap,
interpolation='nearest',
aspect='auto')
plt.pause(pause)
self.fig.canvas.draw()
def plot_generated_batch(i, model,data1,data2):
x_test, y_test = data1
y_pred, x_recon = model.predict([x_test, y_test], batch_size=32)
X_gen = x_recon
X_raw = x_test
Xs1 = to3d(X_raw[:10])
Xg1 = to3d(X_gen[:10])
Xs1 = np.concatenate(Xs1, axis=1)
Xg1 = np.concatenate(Xg1, axis=1)
Xs2 = to3d(X_raw[10:20])
Xg2 = to3d(X_gen[10:20])
Xs2 = np.concatenate(Xs2, axis=1)
Xg2 = np.concatenate(Xg2, axis=1)
x_train, y_train = data2
y_pred, x_recon = model.predict([x_train, y_train], batch_size=32)
X_gen = x_recon
X_raw = x_train
Xs3 = to3d(X_raw[:10])
Xg3 = to3d(X_gen[:10])
Xs3 = np.concatenate(Xs3, axis=1)
Xg3 = np.concatenate(Xg3, axis=1)
Xs4 = to3d(X_raw[10:20])
Xg4 = to3d(X_gen[10:20])
Xs4 = np.concatenate(Xs4, axis=1)
Xg4 = np.concatenate(Xg4, axis=1)
XX = np.concatenate((Xs1,Xg1,Xs2,Xg2,Xs3,Xg3,Xs4,Xg4), axis=0)
plt.imshow(XX)
plt.axis('off')
plt.savefig("./caps_figures/CapsAutoencoder{0:03d}.png".format(i))
plt.pause(3)
plt.close()
def rysuj(self):
try:
parA = float(self.parAedit.text())
parB = float(self.parBedit.text())
parC = float(self.parCedit.text())
if parA == 0 and parB == 0 and parC == 0:
QMessageBox.warning(self, '', "Funkjca tożsamościowa")
return
xlist = []
y = []
import numpy as np
x = np.linspace(-20, 10)
fig, ax = plt.subplots()
ax.axhline(y=0, color='k')
ax.axvline(x=0, color='k')
ax.grid(True, which='both')
#plt.grid(b=None, which='major', axis='both', color='b', linestyle='-', linewidth=2)
for x in self.drange(-10, 10, 0.5):
y.append(parA * x**2 + parB * x + parC)
xlist.append(x)
plt.plot(xlist, y)
plt.pause(0.001)
plt.show()
except ValueError:
QMessageBox.critical(self, "Błąd", "Zły parametr")
return
def main(a):
K = 0 #循环次数控制
algo_choose(a)
init()
fit_list = []
temp = 0
while (True):
for i in range(n):
swarm[i].refresh_memory()
for i in range(n):
swarm[i].refresh_pos()
cal_FitXm()
cal_standard_deviation()
for i in range(n):
group = int(i / (n / M))
#print(group)
swarm[i].escape(group)
update_Td()
print(swarm_best_fitness, swarm_best_pos)
#if (#精度控制条件#):
# break #到达一定精度退出程序
fit_list.append(swarm_best_fitness)
plt.plot(fit_list)
plt.draw()
plt.pause(0.01)
plt.clf()
K = K + 1
temp = swarm_best_fitness
if K == 100: #到达一定次数退出程序
break
def main():
global df
poss_reproduct = 0.1 #繁殖的概率
poss_hybridize = 0.3 #杂交的概率
poss_mutate = 0.6 #变异的概率
for i in range(100000):
te = poss_for_reproduct(df)
r = np.random.random(1)[0]
if (r < poss_reproduct):
df = reproduct(te)
print(1)
if (poss_reproduct < r < poss_reproduct + poss_hybridize):
df = hybridize(te)
print('2')
if (poss_reproduct + poss_hybridize < r):
#pass
df = mutate(te)
print('3')
if (i % 10 == 0):
test = list(df.loc[te.loc[0]['ind']])
plot(test)
plt.draw()
plt.pause(0.01)
plt.clf()
print(te.loc[0][0])
#print(df)
#print(df.loc[te.loc[0]['ind']])
test = list(df.loc[te.loc[0]['ind']])
def histograma(hist):
hist=hist.histogram(255)
## hist.save("hola4Hist.txt")
pylab.plot(hist)
pylab.draw()
pylab.pause(0.0001)
def B_method(p, len):
d = get_distance2(p[0], p[1])
a = 0
b = 1
plt.plot([p[a].x, p[b].x], [p[a].y, p[b].y], color='r')
plt.pause(1)
for i in range(len - 1):
for j in range(i + 1, len):
if i == 0 and j == 1:
continue
distance2 = get_distance2(p[i], p[j])
plt.plot([p[i].x, p[j].x], [p[i].y, p[j].y], color='B', ls='--')
plt.pause(1)
plt.plot([p[i].x, p[j].x], [p[i].y, p[j].y], color='W', alpha=0.99)
plt.pause(1)
if distance2 < d:
plt.plot([p[a].x, p[b].x], [p[a].y, p[b].y],
color='W',
alpha=0.99)
plt.pause(1)
a = i
b = j
d = distance2
plt.plot([p[i].x, p[j].x], [p[i].y, p[j].y], color='r')
plt.pause(1)
print("最近点对为：(%.2f,%.2f) 和 (%.2f,%.2f)，距离为: %.2f" %
(p[a].x, p[a].y, p[b].x, p[b].y, math.sqrt(d)))
def plot(self, bit_stream):
if self.previous_bit_stream != bit_stream.to_list():
self.previous_bit_stream = bit_stream
x = []
y = []
bit = None
for bit_time in bit_stream.to_list():
if bit is None:
x.append(bit_time)
y.append(0)
bit = 0
elif bit == 0:
x.extend([bit_time, bit_time])
y.extend([0, 1])
bit = 1
elif bit == 1:
x.extend([bit_time, bit_time])
y.extend([1, 0])
bit = 0
plt.clf()
plt.plot(x, y)
plt.xlim([0, 10000])
plt.ylim([-0.1, 1.1])
plt.show()
plt.pause(0.005)
def do_next(self, args):
"""next: Shows the next epoch in the animation"""
obs = self.obs
obs.next_epoch()
obs.get_predictions()
obs.show_predictions()
plt.pause(0.000001)
def animate(y, ndim, cmap) :
plt.ion()
if ndim == 5:
plt.figure()
plt.show()
for i in range(y.shape[1]) :
print "Showing batch", i
plt.close('all')
for j in range(y.shape[0]) :
plt.imshow(y[j,i], interpolation='none', cmap=cmap)
plt.pause(0.1)
time.sleep(1)
else:
for i in range(y.shape[1]) :
print "Showing batch", i
plt.close('all')
for j in range(y.shape[0]) :
plt.figure(0)
plt.imshow(y[j,i], interpolation='none', cmap=cmap)
plt.figure(1)
plt.imshow(x[j,i], interpolation='none', cmap=cmap)
plt.pause(0.2)
time.sleep(1)
def return_param_BPM(even_odd, hist_val, num):
if even_odd == "odd":
eo = 1
else:
eo = 0
# print(even_odd,hist_val)
histValues = plt.hist(df[(df['y_vals'] % 2 == eo)
& (df['heights'] < 100)][hist_val],
bins=45)
mphist = midpoints(histValues[1])
meanh = df[(df['y_vals'] % 2 == eo) & (df['heights'] < 100) &
(df['means'] > 0)][hist_val].mean()
stdh = df[(df['y_vals'] % 2 == eo) & (df['heights'] < 100) &
(df['means'] > 0)][hist_val].std()
print(meanh, stdh)
popt, pcov = curve_fit(gaus,
mphist[2:],
histValues[0][2:],
p0=[100, meanh, stdh])
plt.plot(mphist, gaus(mphist, *popt), 'ro:', label='fit')
plt.pause(0.01)
input("wait")
# plt.show()
plt.clf()
return popt[1], popt[2]
def show():
args = get_args()
# Load model
nn.load_parameters(args.model_load_path)
params = nn.get_parameters()
# Show heatmap
for name, param in params.items():
# SSL only on convolution weights
if "conv/W" not in name:
continue
print(name)
n, m, k0, k1 = param.d.shape
w_matrix = param.d.reshape((n, m * k0 * k1))
# Filter x Channel heatmap
fig, ax = plt.subplots()
ax.set_title(
"{} with shape {} \n Filter x (Channel x Heigh x Width)".format(
name, (n, m, k0, k1)))
heatmap = ax.pcolor(w_matrix)
fig.colorbar(heatmap)
plt.pause(0.5)
raw_input("Press Key")
plt.close()
def ani(t):
train_op(t, x, y)
fig.clf()
ax = fig.add_subplot(111)
pts = 300
pylab.plot(x.data.numpy().ravel(), y.data.numpy().ravel(), 'r.')
x_plot = Variable(
torch.linspace(
torch.min(x.data) - 1.,
torch.max(x.data) + 1., pts)[:, None])
y_plot = torch.Tensor(0, 1)
for _ in range(100):
y_plot = torch.cat([y_plot, model(x_plot).data], 1)
y_mu = torch.mean(y_pred, 1).numpy()
y_std = torch.std(y_pred, 1).numpy()
errors = [0.25, 0.39, 0.52, 0.67, 0.84, 1.04, 1.28, 1.64, 2.2]
for e in errors:
ax.fill_between(x_plot.data.numpy().ravel(),
y_mu - e * y_std,
y_mu + e * y_std,
alpha=((3 - e) / 5.5)**1.7,
facecolor='b',
linewidth=1e-3)
pylab.plot(x_plot.data.numpy().ravel(), y_mu, 'k')
ax.set_ylim([torch.min(y.data) - 1, torch.max(y.data) + 1])
ax.set_xlim([torch.min(x.data) - 1, torch.max(x.data) + 1])
pylab.pause(.1)
return ax
def plot(self):
if not self.__first:
self.__fig.clf()
self.__ax.cla()
self.__first = False
self.__fig = plt.figure('residual')
self.__ax = self.__fig.add_subplot(1, 1, 1)
# 現在のパラメータを中心にプロットするためのrangeを計算
self._reset_range()
# 現在のパラメータを履歴に追加
current_param = self.__model.get_param()
self._param_history[0].append(current_param[0])
self._param_history[1].append(current_param[1])
# パラメータを書き換えるのでコピーして参照している他のモジュールへの影響を防ぐ
model = deepcopy(self.__model)
Z = self._compute_residual_distribution(model)
im = self.__ax.pcolormesh(self.__X, self.__Y, Z, cmap='hsv')
self.__ax.plot(self._param_history[0],
self._param_history[1],
marker='.')
self.__fig.colorbar(im)
plt.pause(0.01)
def plot_weight_matrix(rbm_struct, weight_matrix, input_nodes, hidden_nodes):
#mat_to_plot = np.zeros([n_hidden_nodes, n_visible_nodes-1])
sub_mat_val = int(m.sqrt(input_nodes))
total_mat_val = int(m.sqrt(hidden_nodes))
reshaped_mat = np.zeros([hidden_nodes, sub_mat_val, sub_mat_val])
plot_mat = np.zeros(
[total_mat_val, total_mat_val, sub_mat_val, sub_mat_val])
for i in range(0, input_nodes):
a = np.reshape(rbm_struct.weights[i][1][:], (529, 1))
b = np.reshape(a, (23, 23))
mplt.imshow(b, cmap='gray')
mplt.draw()
mplt.pause(0.001)
a = 'weight_matrix' + str(i) + '.png'
mplt.savefig(a)
#for i in range(0, hidden_nodes):
# mat_to_plot = weight_matrix[i][1][:input_nodes]
# mat_to_plot = np.reshape(mat_to_plot, (rbm_struct.h_val, 1))
# reshaped_mat[i][:][:] = np.reshape(mat_to_plot, (sub_mat_val, sub_mat_val))
plot_mat = np.reshape(
reshaped_mat, (total_mat_val, total_mat_val, sub_mat_val, sub_mat_val))
for i in range(0, total_mat_val):
for j in range(0, total_mat_val):
mplt.imshow(plot_mat[i][j], cmap='gray')
mplt.show()
def plotCurrentW(ww, currentData, flush=True):
# Plot the randomly generated data
plt.scatter(currentData[currentData["yp"] == 1]["x1"],
currentData[currentData["yp"] == 1]["x2"],
marker="o")
plt.scatter(currentData[currentData["yp"] == -1]["x1"],
currentData[currentData["yp"] == -1]["x2"],
marker="x")
plt.xlim(-1, 1)
plt.ylim(-1, 1)
plt.xlabel("x1")
plt.ylabel("x2")
# Create data points corresponding the the weights so that they can be plotted on the graph
wDict = {'x1': [], 'x2': []}
leftx2 = (ww[1] - ww[0]) / ww[2]
rightx2 = (-ww[1] - ww[0]) / ww[2]
wDict["x1"].append(-1.0)
wDict["x2"].append(leftx2)
wDict["x1"].append(1.0)
wDict["x2"].append(rightx2)
resultW = pd.DataFrame(wDict)
# Plot the corresponding classification separating line
plt.plot(resultW.x1, resultW.x2)
if flush:
plt.draw()
plt.pause(0.1)
plt.clf()
else:
plt.show()
def draw_graph(self, rnd=None): # pragma: no cover
"""
Draw Graph
"""
plt.clf()
if rnd is not None:
plt.plot(rnd[0], rnd[1], "^k")
for node in self.nodes:
if node.parent is not None:
plt.plot([node.x[0], self.nodes[node.parent].x[0]],
[node.x[1], self.nodes[node.parent].x[1]], "-g")
circles = []
fig = plt.gcf()
ax = fig.gca()
for (ox, oy, size) in self.obstacle_list:
# plt.plot(ox, oy, "ok", ms=30 * size)
circle = plt.Circle((ox, oy), size, fill=False)
circles.append(circle)
p = PatchCollection(circles)
ax.add_collection(p)
ax.set_aspect('equal')
plt.plot(self.start[0], self.start[1], "xr")
plt.plot(self.goal[0], self.goal[1], "xr")
plt.axis([
self.c_space_bounds[0][0], self.c_space_bounds[0][1],
self.c_space_bounds[1][0], self.c_space_bounds[1][1]
])
plt.grid(True)
plt.pause(0.01)
def drawing(x):
plt.figure(num=1)
plt.plot(x, color='g', linewidth=0.8)
plt.title('prob')
plt.grid()
plt.legend(['loss'])
plt.pause(0.0001)
def imshow(inp, title=None):
"""Imshow for Tensor."""
inp = inp.numpy().transpose((1, 2, 0))
plt.imshow(inp)
if title is not None:
plt.title(title)
plt.pause(0.001) # pause a bit so that plots are updated
def ani(t):
train_op(t, x, y)
fig.clf()
ax = fig.add_subplot(111)
pts = 300
pylab.plot(x.data.numpy().ravel(), y.data.numpy().ravel(), 'r.')
x_plot = Variable(
torch.linspace(
torch.min(x.data) - 1.,
torch.max(x.data) + 1., pts)[:, None])
y_plot = np.linspace(
torch.min(y.data) - 1.,
torch.max(y.data) + 1., pts)[:, None]
y_pred = torch.Tensor(0, 1)
for _ in range(100):
y_pred = torch.cat([y_pred, model(x_plot).data], 1)
y_mu = torch.mean(y_pred, 1).numpy()
y_std = torch.std(y_pred, 1).numpy()
kde_skl = KernelDensity(bandwidth=0.5)
grid = []
for i in range(pts):
kde_skl.fit(y_pred.numpy()[i][:, None])
log_pdf = kde_skl.score_samples(y_plot)
grid.append(np.exp(log_pdf)[:, None])
grid = np.asarray(grid).reshape((pts, pts)).T
ax.imshow(grid,
extent=(x_plot.data.numpy().min(),
x_plot.data.numpy().max(), y_plot.max(),
y_plot.min()),
interpolation='bicubic',
cmap=cm.Blues)
ax.set_ylim([torch.min(y.data) - 1, torch.max(y.data) + 1])
ax.set_xlim([torch.min(x.data) - 1, torch.max(x.data) + 1])
pylab.pause(.5)
return ax
def imshow(inp, title=None):
"""Imshow for Tensor."""
import matplotlib.pylab as plt
inp = inp.numpy().transpose((1, 2, 0))
plt.imshow(inp)
if title is not None:
plt.title(title)
plt.pause(0.001) # 갱신이 될 때까지 잠시 기다립니다.
def DisplayData(self,
data,
rec=None,
fut=None,
fig=1,
case_id=0,
output_file=None):
output_file1 = output_file2 = None
if output_file is not None:
name, ext = os.path.splitext(output_file)
output_file1 = '%s_original%s' % (name, ext)
output_file2 = '%s_recon%s' % (name, ext)
data = data[case_id, :].reshape(-1, self.image_size_, self.image_size_)
if rec is not None:
rec = rec[case_id, :].reshape(-1, self.image_size_,
self.image_size_)
enc_seq_length = rec.shape[0]
if fut is not None:
fut = fut[case_id, :].reshape(-1, self.image_size_,
self.image_size_)
if rec is None:
enc_seq_length = self.seq_length_ - fut.shape[0]
else:
assert enc_seq_length == self.seq_length_ - fut.shape[0]
num_rows = 1
plt.figure(2 * fig - 1, figsize=(20, 1))
plt.clf()
for i in xrange(self.seq_length_):
plt.subplot(num_rows, self.seq_length_, i + 1)
plt.imshow(data[i, :, :],
cmap=plt.cm.gray,
interpolation="nearest")
plt.axis('off')
plt.draw()
if output_file1 is not None:
print output_file1
plt.savefig(output_file1, bbox_inches='tight')
plt.figure(2 * fig, figsize=(20, 1))
plt.clf()
for i in xrange(self.seq_length_):
if rec is not None and i < enc_seq_length:
plt.subplot(num_rows, self.seq_length_, i + 1)
plt.imshow(rec[rec.shape[0] - i - 1, :, :],
cmap=plt.cm.gray,
interpolation="nearest")
if fut is not None and i >= enc_seq_length:
plt.subplot(num_rows, self.seq_length_, i + 1)
plt.imshow(fut[i - enc_seq_length, :, :],
cmap=plt.cm.gray,
interpolation="nearest")
plt.axis('off')
plt.draw()
if output_file2 is not None:
print output_file2
plt.savefig(output_file2, bbox_inches='tight')
else:
plt.pause(0.1)
def plot_data():
datalist = [ ( plt.loadtxt("myfile.txt"), "label")]
for data, label in datalist:
plt.plot( data[:,0], data[:,1], label=label )
plt.legend()
plt.title("Plot of coupled oscillator")
plt.xlabel("Time")
plt.ylabel("Position")
plt.show()
plt.pause(0.000005)
def sample02(fnum):
# インタラクティブモード
x = np.mat(np.r_[0:1:0.01])
y2 = np.power(x,2)
pl.ion() #非インタラクティブにする
pl.figure(fnum)
pl.plot(x.A1, y2.A1) #これは成功
pl.grid(True)
# pl.show() #インタラクティブモードだと、showは使わない
pl.draw() #drawを使う
print "end2"
# time.sleep(1)
pl.pause(1)
def plot(self, title='', confidence_interval=DEFAULT_CONFIDENCE_INTERVAL, block=False, pause=0, figsize=None):
self.init_dataframe_errors(confidence_interval)
assert isinstance(self, Results)
if not self.periods:
raise ValueError("Results have no periods to plot")
fig, axes = plt.subplots(nrows=3, ncols=1, figsize=figsize)
if title:
fig.canvas.set_window_title(title)
if isinstance(self.observation_date, datetime.datetime):
observation_date = self.observation_date.date()
else:
observation_date = self.observation_date
fig.suptitle('Obs {}, rate {}%, path {}, pert {}, conf {}%'.format(
observation_date, self.interest_rate, self.path_count,
self.perturbation_factor,
confidence_interval))
with pandas.plotting.plot_params.use('x_compat', False):
# Todo: Try to get the box plots working:
# https://stackoverflow.com/questions/38120688/pandas-box-plot-for-multiple-column
# if len(results.prices_raw) == 1:
# prices = results.prices_raw[0]
# seaborn.boxplot(prices, prices.to_series().apply(lambda x: x.strftime('%Y%m%d')), ax=axes[0])
self.prices_mean.plot(ax=axes[0], kind='bar', yerr=self.prices_errors)
axes[0].set_title('Market prices')
axes[0].get_xaxis().set_visible(False)
self.hedges_mean.plot(ax=axes[1], kind='bar', yerr=self.hedges_errors).axhline(0, color='0.5')
axes[1].set_title('Delta hedges')
axes[1].get_xaxis().set_visible(False)
self.cash_mean.plot(ax=axes[2], kind='bar', yerr=self.cash_errors, color='g').axhline(0, color='0.5')
axes[2].set_title('Cash account')
axes[2].get_xaxis().set_visible(True)
fig.autofmt_xdate(rotation=30)
if pause and not block:
plt.pause(pause)
else:
plt.show(block=block)
def animate_signals(nb_signals, incre, fs=256, refresh_rate=30., anim_dur=10., figsize=(12,8)):
"""
Draw and update a figure in real-time representing the summation of many
sine waves, to explain the concept of Fourier decomposition.
INPUTS
nb_signals : number of signals to sum together
incre : increment, in Hz, between each of the signals
fs : sampling frequency
refresh_rate : refresh rate of the animation
anim_dur : approximate duration of the animation, in seconds
"""
# Initialize values that remain constant throughout the animation
A = 1
t = np.linspace(0, 2, fs)
offsets = np.arange(nb_signals+1).reshape((nb_signals+1,1))*(A*(nb_signals+1))
freqs = np.arange(nb_signals)*incre
# Initialize the figure
fig, ax = plt.subplots(figsize=figsize)
ax.hold(True)
plt.xlabel('Time')
ax.yaxis.set_ticks(offsets)
ax.set_yticklabels([str(f)+' Hz' for f in freqs] + ['Sum'])
ax.xaxis.set_ticks([])
# Initialize the Line2D elements for each signal
sines = np.array([sinewave(A, f, 0, t) for f in freqs])
sines = np.vstack((sines, np.sum(sines, axis=0))) + offsets
points = [ax.plot(t, x)[0] for x in sines]
# Animation refresh loop
for i in np.arange(anim_dur*refresh_rate):
# Update time
t = np.linspace(0, 2, fs) + i*fs/refresh_rate
# Update signals
sines = np.array([sinewave(A, f, 0, t) for f in freqs])
sines = np.vstack((sines, np.sum(sines, axis=0))) + offsets
# Update figure
for p, x in zip(points, sines):
p.set_ydata(x)
# Wait before starting another cycle
plt.pause(1./refresh_rate)
def cruiseControllerNoPID(n, req_speed, initial_speed = 0, accelerate_step = 1):
speeds = []
current_speed = initial_speed
ns = []
x = 0
acc = []
while x < n + 1:
x += 1
#print current_speed
#plt.axis([-10, n + 10 , -10, req_speed + 10])
plt.axis([-10, n + 10 , -1.5, 0.75])
speeds.append(current_speed)
acc.append(speeds[x - 1] - speeds[x - 2])
ns.append(x)
#plt.plot(ns, speeds, 'g-',label = "speed")
plt.plot(ns, acc, 'r-', label = "input (force applied to pedal)")
if (x == 1):
plt.legend()
plt.xlabel("time")
plt.ylabel("Input (Force applied to pedal)")
plt.pause(0.000001)
if (current_speed < req_speed):
#accelerate
current_speed += accelerate_step
elif (current_speed > req_speed):
#apply brakes
current_speed -= accelerate_step
else:
#speed reduction when pedal isn't pressed
current_speed -= 1
if (current_speed < 0):
current_speed = 0
def PIDController(n, req_speed, initial_speed = 0, accelerate_step = 1):
speeds = []
current_speed = initial_speed
ns = []
x = 0
acc = []
req_speeds = []
accumulated_e = 0
while x < n + 1:
x += 1
#print current_speed
plt.axis([-10, n + 10 , -10, req_speed + 10])
#plt.axis([-10, n + 10 , -1.5, 0.75])
speeds.append(current_speed)
req_speeds.append(req_speed)
acc.append(speeds[x - 1] - speeds[x - 2])
ns.append(x)
plt.plot(ns, speeds, 'g-',label = "speed")
plt.plot(ns, req_speeds, 'r--', label = "required speed")
#plt.plot(ns, acc, 'r-', label = "input (force applied to pedal)")
if (x == 1):
plt.legend()
plt.xlabel("time")
plt.ylabel("Car's speed")
plt.pause(0.000001)
drag = 0.03
e = req_speed - current_speed
KI = 0.07
Kd = 0.1
Kp = 0.1
accumulated_e += e
diff_e = speeds[x - 1] - speeds[x - 2]
current_speed += (e * Kp - drag * current_speed) + (Kp * accumulated_e) + (Kd * diff_e)
print current_speed
def DisplayData(self, data, rec=None, fut=None, fig=1, case_id=0, output_file=None):
output_file1 = output_file2 = None
if output_file is not None:
name, ext = os.path.splitext(output_file)
output_file1 = '%s_original%s' % (name, ext)
output_file2 = '%s_recon%s' % (name, ext)
data = data[case_id, :].reshape(-1, self.image_size_, self.image_size_)
if rec is not None:
rec = rec[case_id, :].reshape(-1, self.image_size_, self.image_size_)
enc_seq_length = rec.shape[0]
if fut is not None:
fut = fut[case_id, :].reshape(-1, self.image_size_, self.image_size_)
if rec is None:
enc_seq_length = self.seq_length_ - fut.shape[0]
else:
assert enc_seq_length == self.seq_length_ - fut.shape[0]
num_rows = 1
plt.figure(2*fig-1, figsize=(20, 1))
plt.clf()
for i in xrange(self.seq_length_):
plt.subplot(num_rows, self.seq_length_, i+1)
plt.imshow(data[i, :, :], cmap=plt.cm.gray, interpolation="nearest")
plt.axis('off')
plt.draw()
if output_file1 is not None:
print output_file1
plt.savefig(output_file1, bbox_inches='tight')
plt.figure(2*fig, figsize=(20, 1))
plt.clf()
for i in xrange(self.seq_length_):
if rec is not None and i < enc_seq_length:
plt.subplot(num_rows, self.seq_length_, i + 1)
plt.imshow(rec[rec.shape[0] - i - 1, :, :], cmap=plt.cm.gray, interpolation="nearest")
if fut is not None and i >= enc_seq_length:
plt.subplot(num_rows, self.seq_length_, i + 1)
plt.imshow(fut[i - enc_seq_length, :, :], cmap=plt.cm.gray, interpolation="nearest")
plt.axis('off')
plt.draw()
if output_file2 is not None:
print output_file2
plt.savefig(output_file2, bbox_inches='tight')
else:
plt.pause(0.1)
def plot(self, cmap='Greys',pause=1): if self.fig is None : self.fig = plt.figure(1,figsize=(24,5)) self.ax1 = self.fig.add_subplot(211) self.ax2 = self.fig.add_subplot(212) plt.tight_layout() #plt.show() else : plt.figure(1) #... build numpy array... FOR EXAMPLE : aview = np.random.randint(0,2,(5,300)) pview = np.random.randint(0,2,(5,300)) self.ax1.imshow(aview, cmap=cmap, interpolation='nearest', aspect='auto') self.ax2.imshow(pview, cmap=cmap, interpolation='nearest', aspect='auto') plt.pause(pause) self.fig.canvas.draw()
def showImg (self):
if self._pltmoon == 1:
# Convert UTC unix time to datetime
self._obssite.date = datetime.datetime.fromtimestamp(self._im._tobs);
# Compute azi/alt
self._moon.compute(self._obssite);
self._casa.compute(self._obssite);
if self._moon.alt < 0:
print 'Moon below horizon at time ', _obssite.date;
else:
# Compute l,m of moon's position, in units of array indices
moon_l = -(numpy.cos(self._moon.alt) * numpy.sin(self._moon.az));
moon_m = (numpy.cos(self._moon.alt) * numpy.cos(self._moon.az));
# print 'l/m: %f, %f' % (moon_l, moon_m);
moon_l = moon_l/self._im._dl + self._im._npix/2;
moon_m = moon_m/self._im._dl + self._im._npix/2;
# print 'Moon: RA/dec = %f/%f, alt/az = %f/%f, lind/mind = %f/%f, dl=%f'% (self._moon.ra, self._moon.dec, self._moon.alt, self._moon.az, moon_l, moon_m, self._im._dl);
casa_l = -(numpy.cos(self._casa.alt) * numpy.sin(self._casa.az));
casa_m = (numpy.cos(self._casa.alt) * numpy.cos(self._casa.az));
# print 'l/m: %f, %f' % (casa_l, casa_m);
casa_l = casa_l/self._im._dl + self._im._npix/2;
casa_m = casa_m/self._im._dl + self._im._npix/2;
# print 'CasA: RA/dec = %f/%f, alt/az = %f/%f, lind/mind = %f/%f, dl=%f'% (self._casa.ra, self._casa.dec, self._casa.alt, self._casa.az, casa_l, casa_m, self._im._dl);
# Create a circle in the skymap centered at the location of the moon.
# blacking out a couple of pixels for now.
self._im._skymap[moon_m-1:moon_m+1, moon_l-1:moon_l+1] = 5e10;
self._im._skymap[casa_m-1:casa_m+1, casa_l-1:casa_l+1] = 5e10;
self._imgplt.set_data (abs(self._im._skymap[:,:]));
plt.title ('XX - %f' % self._im._tobs);
plt.draw();
plt.pause(0.001);
if self._wrpng == 1:
plt.savefig ('%s/%.0f_XX.png' % (self._fprefix,self._im._tobs));
def classify(im, K, show=False):
""" classify - classify an image using k-means algorithm.
Arguments:
* im - an image (ndarray) or a filename.
* K - number of classes.
* show = False - if True, display the image and the classes.
Return:
* imClasses - the classes of the image (ndarray, 2D).
See also:
* kMeans - K-means algorithm.
"""
a = Image(im)
a = a.histogram(255)
# Load image
if type(im) is str:
im = plt.imread(im)
# Reshape and apply k-mean
values = im.reshape((im.shape[0] * im.shape[1], -1))
centers, classes = kMeans(values, K)
imClasses = classes.reshape((im.shape[0], im.shape[1]))
## imClasses.save("imClassesH.png")
## imClassesHl=image("imClassesH.png")
## imClassHisto=imClassesH1.histogram(255)
# Display
if show:
cmap = 'gray' if im.ndim < 3 else None
plt.subplot(221), plt.imshow(im, cmap=cmap), plt.title('Imagen Gray')
plt.subplot(222), plt.imshow(imClasses), plt.title('Segmentacion por Clases')
pylab.subplot(223), pylab.plot(a), pylab.draw(), pylab.pause(0.0001), pylab.title('Histograma Img Gray')
pylab.subplot(224), pylab.plot(a), pylab.draw(), pylab.pause(0.0001), pylab.title('Histograma Img Segmentada')
plt.show()
# The end
return imClasses
def showImg(self):
if self._pltmoon == 1:
# Convert UTC unix time to datetime
self._obssite.date = datetime.datetime.fromtimestamp(self._im[0]._tobs)
# Compute azi/alt
self._moon.compute(self._obssite)
self._casa.compute(self._obssite)
if self._moon.alt < 0:
print "Moon below horizon at time ", self._obssite.date
else:
# Compute l,m of moon's position, in units of array indices
moon_l = -(numpy.cos(self._moon.alt) * numpy.sin(self._moon.az))
moon_m = numpy.cos(self._moon.alt) * numpy.cos(self._moon.az)
print "moon l/m: %f, %f" % (moon_l, moon_m)
# moon_l = moon_l/self._im[0]._dl + self._im[0]._npix/2;
# moon_m = moon_m/self._im[0]._dl + self._im[0]._npix/2;
# print 'Moon: RA/dec = %f/%f, alt/az = %f/%f, lind/mind = %f/%f, dl=%f'% (self._moon.ra, self._moon.dec, self._moon.alt, self._moon.az, moon_l, moon_m, self._im._dl);
if self._casa.alt < 0:
print "CasA below horizon at time ", self._obssite.date
else:
casa_l = -(numpy.cos(self._casa.alt) * numpy.sin(self._casa.az))
casa_m = numpy.cos(self._casa.alt) * numpy.cos(self._casa.az)
print "CasA l/m: %f, %f" % (casa_l, casa_m)
# casa_l = casa_l/self._im[0]._dl + self._im[0]._npix/2;
# casa_m = casa_m/self._im[0]._dl + self._im[0]._npix/2;
# print 'CasA: RA/dec = %f/%f, alt/az = %f/%f, lind/mind = %f/%f, dl=%f'% (self._casa.ra, self._casa.dec, self._casa.alt, self._casa.az, casa_l, casa_m, self._im._dl);
self._imgplt.set_data(abs(self._im[0]._skymap[:, :]))
plt.title("%s - %f" % (self._im[0]._pol, self._im[0]._tobs))
# if self._moon.alt > 0:
# self._im[0]._skymap[moon_m-1:moon_m+1, moon_l-1:moon_l+1] = 5e10;
# plt.annotate ('*',xy=(-moon_l,-moon_m),color='yellow');
# plt.annotate ('*',xy=(-casa_m,casa_l),color='white');
plt.draw()
plt.pause(0.001)
if self._wrpng == 1:
plt.savefig("%s/%.0f_XX.png" % (self._fprefix, self._im[0]._tobs))
def run(self):
fig = plt.figure()
fig.subplots_adjust(top=0.8)
ax2 = fig.add_axes([0.1, 0.1, 0.8, 0.8])
ax2.set_xlabel('Simulation of boat')
(ax_min_x, ax_min_y, axis_len) = (-20, -20, 40)
while(self.run_th):
self.lock.acquire()
th_data = copy.deepcopy(temp_data)
self.lock.release()
self.run_th = th_data.run
plt.cla() # Clear axis
plt.clf() # Clear figure
fig.subplots_adjust(top=0.8)
ax2 = fig.add_axes([0.1, 0.1, 0.8, 0.8])
ax2.set_xlabel('Simulation of boat %0.1f %0.1f speed:%0.1f m/s rudder:%0.3f\
lat %.5f long %.5f ' %
(wrapTo2Pi(-th_data.theta+np.pi/2)*180/np.pi,
wrapTo2Pi(-th_data.phi+np.pi/2)*180/np.pi,
th_data.speed,
th_data.delta_r,
th_data.latitude, th_data.longitude))
cds.draw_boat(ax2, 1, th_data.x, th_data.y,
th_data.theta, th_data.delta_r, th_data.delta_s)
ax_min_x = x-axis_len/2.0
ax_min_y = y-axis_len/2.0
cds.draw_wind_direction(ax2, (ax_min_x+1, ax_min_y+1), axis_len, 1, th_data.phi)
plt.axis([ax_min_x, ax_min_x+axis_len,
ax_min_y, ax_min_y+axis_len])
plt.draw()
plt.pause(0.001)
print("Stopping Draw Thread")
plt.close()
fig0, ax0 = plt.subplots(num=0)
plt.title("COOLFluiD Convergence Monitoring")
if res=="T":
lh0, = ax0.plot(iterations,T_res,"+",color="#1f77b4")
ax0.set_ylabel(r"$\log\, T$ Residual",color="#1f77b4")
elif res=="rho":
lh0, = ax0.plot(iterations,rho_res,"+",color="#1f77b4")
ax0.set_ylabel(r"$\log\, \varrho$ Residual",color="#1f77b4")
ax0.set_xlabel(r"Number of iterations")
ax0.tick_params("y",colors="#1f77b4")
ax0.grid(True)
ax1 = ax0.twinx()
lh1, = ax1.plot(iterations,CFL,".",color="#d62728")
ax1.set_ylabel(r"CFL",color="#d62728")
ax1.tick_params("y",colors="#d62728")
plt.pause(1) # flush the figure
plt.close(1)
fig1, ax2 = plt.subplots(num=1)
plt.title("COOLFluiD Convergence Monitoring")
lh2, = ax2.plot(iterations,Bx_res,"+",color="#ff7f0e",label=r"$\log\,B$ residual")
lh3, = ax2.plot(iterations,Ex_res,"+",color="#2ca02c",label=r"$\log\,E$ residual")
lh4, = ax2.plot(iterations,vx_res,"+",color="#7f7f7f",label=r"$\log\,U$ residual")
if res=="T":
# Plot the residual that is not already monitored in Fig. 1
lh5, = ax2.plot(iterations,rho_res,"+",color="#9467bd",label=r"$\log\,\varrho$ residual")
elif res=="rho":
lh5, = ax2.plot(iterations,T_res,"+",color="#9467bd",label=r"$\log\,T$ residual")
plt.legend(loc=1)
ax2.set_xlabel(r"Number of iterations")
ax2.tick_params("y",colors="k")
def main( conf_file='config.cfg', logfile=None ):
#%% parameters
print "reading config parameters..."
config, pars = front_end.parser( conf_file )
if pars.has_key('logging') and pars['logging']:
print "recording configuration file..."
front_end.record_config_file( pars )
logfile = front_end.make_logfile_name( pars )
#%% create and initialize the network
if pars['train_load_net'] and os.path.exists(pars['train_load_net']):
print "loading network..."
net = netio.load_network( pars )
# load existing learning curve
lc = zstatistics.CLearnCurve( pars['train_load_net'] )
else:
if pars['train_seed_net'] and os.path.exists(pars['train_seed_net']):
print "seeding network..."
net = netio.load_network( pars, is_seed=True )
else:
print "initializing network..."
net = netio.init_network( pars )
# initalize a learning curve
lc = zstatistics.CLearnCurve()
# show field of view
print "field of view: ", net.get_fov()
print "output volume info: ", net.get_outputs_setsz()
# set some parameters
print 'setting up the network...'
vn = utils.get_total_num(net.get_outputs_setsz())
eta = pars['eta'] #/ vn
net.set_eta( eta )
net.set_momentum( pars['momentum'] )
net.set_weight_decay( pars['weight_decay'] )
# initialize samples
outsz = pars['train_outsz']
print "\n\ncreate train samples..."
smp_trn = front_end.CSamples(config, pars, pars['train_range'], net, outsz, logfile)
print "\n\ncreate test samples..."
smp_tst = front_end.CSamples(config, pars, pars['test_range'], net, outsz, logfile)
# initialization
elapsed = 0
err = 0
cls = 0
# interactive visualization
plt.ion()
plt.show()
# the last iteration we want to continue training
iter_last = lc.get_last_it()
print "start training..."
start = time.time()
print "start from ", iter_last+1
for i in xrange(iter_last+1, pars['Max_iter']+1):
vol_ins, lbl_outs, msks = smp_trn.get_random_sample()
# forward pass
vol_ins = utils.make_continuous(vol_ins, dtype=pars['dtype'])
props = net.forward( vol_ins )
# cost function and accumulate errors
props, cerr, grdts = pars['cost_fn']( props, lbl_outs )
err = err + cerr
cls = cls + cost_fn.get_cls(props, lbl_outs)
# mask process the gradient
grdts = utils.dict_mul(grdts, msks)
# run backward pass
grdts = utils.make_continuous(grdts, dtype=pars['dtype'])
net.backward( grdts )
if pars['is_malis'] :
malis_weights = cost_fn.malis_weight(props, lbl_outs)
grdts = utils.dict_mul(grdts, malis_weights)
if i%pars['Num_iter_per_test']==0:
# test the net
lc = test.znn_test(net, pars, smp_tst, vn, i, lc)
if i%pars['Num_iter_per_show']==0:
# anneal factor
eta = eta * pars['anneal_factor']
net.set_eta(eta)
# normalize
err = err / vn / pars['Num_iter_per_show']
cls = cls / vn / pars['Num_iter_per_show']
lc.append_train(i, err, cls)
# time
elapsed = time.time() - start
elapsed = elapsed / pars['Num_iter_per_show']
show_string = "iteration %d, err: %.3f, cls: %.3f, elapsed: %.1f s/iter, learning rate: %.6f"\
%(i, err, cls, elapsed, eta )
if pars.has_key('logging') and pars['logging']:
utils.write_to_log(logfile, show_string)
print show_string
if pars['is_visual']:
# show results To-do: run in a separate thread
front_end.inter_show(start, lc, eta, vol_ins, props, lbl_outs, grdts, pars)
if pars['is_rebalance'] and 'aff' not in pars['out_type']:
plt.subplot(247)
plt.imshow(msks.values()[0][0,0,:,:], interpolation='nearest', cmap='gray')
plt.xlabel('rebalance weight')
if pars['is_malis']:
plt.subplot(248)
plt.imshow(malis_weights.values()[0][0,0,:,:], interpolation='nearest', cmap='gray')
plt.xlabel('malis weight (log)')
plt.pause(2)
plt.show()
# reset err and cls
err = 0
cls = 0
# reset time
start = time.time()
if i%pars['Num_iter_per_save']==0:
# save network
netio.save_network(net, pars['train_save_net'], num_iters=i)
lc.save( pars, elapsed )
def move(self):
""" Move the Person
"""
if self.pdshow:
fig =plt.gcf()
fig,ax=self.L.showG('w',labels=False,alphacy=0.,edges=False,fig=fig)
plt.draw()
plt.ion()
while True:
if self.moving:
if self.sim.verbose:
print 'meca: updt ag ' + self.ID + ' @ ',self.sim.now()
# if np.allclose(conv_vecarr(self.destination)[:2],self.L.Gw.pos[47]):
# import ipdb
# ipdb.set_trace()
while self.cancelled:
yield passivate, self
print "Person.move: activated after being cancelled"
checked = []
for zone in self.world.zones(self):
if zone not in checked:
checked.append(zone)
zone(self)
# updating acceleration
acceleration = self.steering_mind(self)
acceleration = acceleration.truncate(self.max_acceleration)
self.acceleration = acceleration
# updating velocity
velocity = self.velocity + acceleration * self.interval
self.velocity = velocity.truncate(self.max_speed)
if velocity.length() > 0.2:
# record direction only when we've really had some
self.localy = velocity.normalize()
self.localx = vec3(self.localy.y, -self.localy.x)
# updating position
self.position = self.position + self.velocity * self.interval
# self.update()
self.position.z=0
self.world.update_boid(self)
self.net.update_pos(self.ID,conv_vecarr(self.position),self.sim.now())
p=conv_vecarr(self.position).reshape(3,1)
v=conv_vecarr(self.velocity).reshape(3,1)
a=conv_vecarr(self.acceleration).reshape(3,1)
# fill panda dataframe 2D trajectory
self.df = self.df.append(pd.DataFrame({'t':pd.Timestamp(self.sim.now(),unit='s'),
'x':p[0],
'y':p[1],
'vx':v[0],
'vy':v[1],
'ax':a[0],
'ay':a[1]},
columns=['t','x','y','vx','vy','ax','ay']))
if self.pdshow:
ptmp =np.array([p[:2,0],p[:2,0]+v[:2,0]])
if hasattr(self, 'pl'):
self.pl[0].set_data(self.df['x'].tail(1),self.df['y'].tail(1))
self.pla[0].set_data(ptmp[:,0],ptmp[:,1])
else :
self.pl = ax.plot(self.df['x'].tail(1),self.df['y'].tail(1),'o',color=self.color,ms=self.radius*10)
self.pla = ax.plot(ptmp[:,0],ptmp[:,1],'r')
# try:
# fig,ax=plu.displot(p[:2],p[:2]+v[:2],'r')
# except:
# pass
# import ipdb
# ipdb.set_trace()
plt.draw()
plt.pause(0.0001)
if 'mysql' in self.save:
self.db.writemeca(self.ID,self.sim.now(),p,v,a)
if 'txt' in self.save:
pyu.writemeca(self.ID,self.sim.now(),p,v,a)
# new target when arrived in poi
if self.arrived and\
(self.L.pt2ro(self.position) ==\
self.L.Gw.node[self.rooms[1]]['room']):
self.arrived = False
if self.endpoint:
self.endpoint=False
self.roomId = self.nextroomId
# remove the remaining waypoint which correspond
# to current room position
del self.waypoints[0]
del self.rooms[0]
# del self.dlist[0]
#
# If door lets continue
#
#
# ig destination --> next room
#
#adjroom = self.L.Gr.neighbors(self.roomId)
#Nadjroom = len(adjroom)
if self.cdest == 'random':
# self.nextroomId = int(np.floor(random.uniform(0,self.L.Gr.size())))
self.nextroomId = random.sample(self.L.Gr.nodes(),1)[0]
# test 1 ) next != actualroom
# 2 ) nextroom != fordiden room
# 3 ) room not share without another agent
while self.nextroomId == self.roomId or (self.nextroomId in self.forbidroomId):# or (self.nextroomId in self.sim.roomlist):
# self.nextroomId = int(np.floor(random.uniform(0,self.L.Gr.size())))
self.nextroomId = random.sample(self.L.Gr.nodes(),1)[0]
elif self.cdest == 'file':
self.room_counter=self.room_counter+1
if self.room_counter >= self.nb_room:
self.room_counter=0
self.nextroomId=self.room_seq[self.room_counter]
self.wait=self.room_wait[self.room_counter]
#self.sim.roomlist.append(self.nextroomId) # list of all destiantion of all nodes in object sim
self.rooms, wp = self.L.waypointGw(self.roomId,self.nextroomId)
# self.dlist = [i in self.L.Gw.ldo for i in self.rooms]
for tup in wp[1:]:
self.waypoints.append(vec3(tup))
#nextroom = adjroom[k]
# print "room : ",self.roomId
# print "nextroom : ",self.nextroomId
#p_nextroom = self.L.Gr.pos[self.nextroomId]
#setdoors1 = self.L.Gr.node[self.roomId]['doors']
#setdoors2 = self.L.Gr.node[nextroom]['doors']
#doorId = np.intersect1d(setdoors1,setdoors2)[0]
#
# coord door
#
#unode = self.L.Gs.neighbors(doorId)
#p1 = self.L.Gs.pos[unode[0]]
#p2 = self.L.Gs.pos[unode[1]]
#print p1
#print p2
#pdoor = (np.array(p1)+np.array(p2))/2
self.destination = self.waypoints[0]
if self.sim.verbose:
print 'meca: ag ' + self.ID + ' wait ' + str(self.wait)#*self.interval)
yield hold, self, self.wait
else:
del self.waypoints[0]
del self.rooms[0]
# del self.dlist[0]
#print "wp : ", self.waypoints
if len(self.waypoints)==1:
self.endpoint=True
self.destination = self.waypoints[0]
#print "dest : ", self.destination
else:
yield hold, self, self.interval
else:
# self.update()
self.world.update_boid(self)
self.net.update_pos(self.ID,conv_vecarr(self.position),self.sim.now())
yield hold, self, self.interval
f.build_operators()
f.regrid() ; f.solve_steady()
tt = interp1d(f.z,f.T,kind="cubic")
Tst = np.append(Tst,tt(f.z_ref))
Xst = np.append(Xst,f.chi_ref)
G = np.append(G,np.trapz(f.d*f.G,f.z))
S = np.append(S,np.trapz(f.d*f.S,f.z))
if len(Tst)>1:
dT = abs(Tst[-1]-Tst[-2])
y2 = y1 ; y1 = y0 ; y0 = f.Y
z2 = z1 ; z1 = z0 ; z0 = f.z
pl.subplot(3,1,1) ; pl.plot(f.z,f.T)
pl.subplot(3,1,2) ; pl.plot(f.z,f.Y)
pl.subplot(3,1,3) ; pl.plot(f.z,f.s)
pl.pause(.1) ; pl.clf()
while Tst[-1] > Tmix+deltaT:
w0,w1,w2 = quadwts(Tst[-1]-deltaT,Tst[-1],Tst[-2],Tst[-3])
f.chi_ref = np.exp(w0*np.log(Xst[-1])+w1*np.log(Xst[-2])+w2*np.log(Xst[-3]))
f.build_operators()
#yy1 = interp1d(z1,y1,axis=0)
#yy2 = interp1d(z2,y2,axis=0)
#f.Y = w0*y0+w1*yy1(z0)+w2*yy2(z0)
f.regrid() ; f.solve_steady()
tt = interp1d(f.z,f.T,kind="cubic")
Tst = np.append(Tst,tt(f.z_ref))
Xst = np.append(Xst,f.chi_ref)
G = np.append(G,np.trapz(f.d*f.G,f.z))
for i in range(0,999):
gc=0
for j in range(0,len(dnalist)):
dnalist[j]=(dnalist[j][0]+'{:02d}'.format(j),mutated_string(dnalist[j][1][:]))
gc+=gc_content(dnalist[j][1])
gccont=gc/len(dnalist)
numseqs,dnalist = sample_seqs(dnalist)
x.append(i)
y.append(gccont)
y2.append(numseqs)
print(numseqs,gccont, sep='\t')
line.set_xdata(np.array(x))
line.set_ydata(np.array(y))
line2.set_xdata(np.array(x))
line2.set_ydata(np.array(y2))
if (i>500):
plt.subplot(211)
plt.xlim(i-500,i+10)
if (i>100):
plt.subplot(212)
plt.xlim(i-100,i+10)
plt.draw()
plt.pause(0.005)
for i in dnalist:
#print('>'+i[0],''.join(i[1]),sep='\n')
pass
# of Life
if __name__ == '__main__':
N = 100.0
domain = np.zeros((N,N), dtype=np.int)
domain = add_glider(domain,(10,50))
domain = add_smallgrower(domain,(70,70))
domain = add_gospergun(domain,(20,20))
# note infinate loop
# exits with ctrl-c
while True:
plt.imshow(domain)
#plt.show()
plt.pause(.1)
plt.draw()
plt.clf()
domain = update_array(domain)
def plot_rssi(self, port):
self.count = 0
self.dt = 0.5
self.tlen = 120
# Generate mesh for plotting
Y, X = np.mgrid[slice(0 - .5, 26 + 1.5, 1), slice(0 - self.dt / 2, self.tlen + 1 - self.dt / 2, self.dt)]
Z = np.zeros_like(X)
# X and Y are bounds, so Z should be the value *inside* those bounds.
# Therefore, remove the last value from the Z array.
Z = Z[:, :-1]
logging.debug("Creating figure")
plt.figure()
pcm = plt.pcolormesh(X, Y, Z, vmin=-128, vmax=0, cmap=plt.cm.get_cmap('jet'))
plt.xlim([0, self.tlen])
plt.ylim([0, 26])
plt.colorbar(label="Measured signal level [dB]")
plt.ylabel("Channel number")
plt.xlabel("Time [s]")
plt.ion()
logging.debug("Show plot window")
plt.show()
ch_min = 26
ch_max = 0
last_update = time.time()
logging.info("Begin collecting data from serial port")
while True:
line = port.readline().rstrip()
pkt_data = re.match(r"\[([-+]?\d+),\s*([-+]?\d+),\s*([-+]?\d+)\]\s*(.*)", line.decode(errors='replace'))
if pkt_data:
now = time.time()
try:
iface_id = int(pkt_data.group(1))
timestamp = int(pkt_data.group(2))
count = int(pkt_data.group(3))
tidx = int(timestamp / (self.dt * 1000000)) % (Z.shape[1])
except ValueError:
continue
logging.debug("data: tidx=%d if=%d cnt=%d t=%d", tidx, iface_id, count, timestamp)
raw = pkt_data.group(4)
resize = False
for ch_ed in raw.split(","):
try:
pair = ch_ed.split(":")
ch = int(pair[0])
ed = float(pair[1])
if ch < ch_min:
ch_min = ch
resize = True
if ch > ch_max:
ch_max = ch
resize = True
Z[ch, tidx] = ed
except (ValueError, IndexError):
continue
if resize:
logging.debug("resize: %d %d" % (ch_min, ch_max))
plt.ylim([ch_min - .5, ch_max + .5])
if now > last_update + 1:
last_update = now
pcm.set_array(Z.ravel())
pcm.autoscale()
pcm.changed()
plt.pause(0.01)
import signal;
import vismon;
def sighdlr (signal, frame):
print 'Keyboard interrupt! aborting...'
sys.exit(0);
if __name__ == "__main__":
rec = 0;
fid = open (sys.argv[1], 'r');
im = vismon.imager ();
im.readMetaFromFile (fid);
tobs, fobs, img = im.readImgFromFile (fid);
print 'Rec: %04d, Time: %.0f, Freq: %.0f' % (rec, tobs, fobs);
img.shape = (im._npix, im._npix);
imgplt = plt.imshow (abs(img));
plt.colorbar();
signal.signal(signal.SIGINT, sighdlr);
while True:
tobs, fobs, img = im.readImgFromFile (fid);
print 'Rec: %d, Time: %.0f, Freq: %.0f' % (rec, tobs, fobs);
img.shape = (im._npix, im._npix);
imgplt.set_data (abs(img));
plt.draw();
plt.title ('XX - %f' % tobs);
plt.pause(0.01);
# plt.clf();
rec = rec + 1;
if __name__ == '__main__':
# Plotting distance modulus as a function of redshift for all C11 parameters
types = ["C11_Combined", "C11_SALT2_stat_sys", "C11_SiFTO_stat_sys", "C11_SALT2_stat", "C11_SiFTO_stat", "C11_reanalized", "C11_recalibrated", "JLA"]
dirname = "/Users/anita/Documents/Grad_First_Year/Summer_2016_research/Codes/"
cvals = ["black","blue","red","green","yellow","orange","magenta","teal"]
plt.ion()
for i in range(len(types)):
zcmb, mu= dist_modul(dirname,"table3_new.txt",ptype=types[i])
plt.plot(zcmb, mu,"o", ms=4, color=cvals[i], label=str(types[i]))
plt.title(types[i])
plt.pause(1.5)
plt.draw()
plt.xlabel("$z_{cmb}$", fontsize=25)
plt.ylabel(r"$\mu = m_B - (M_B - \alpha X_1 + \beta C)$", fontsize=20)
plt.legend(loc='best')
plt.figure(figsize=(9.5, 8.5))
# load up a single event, so we can get the subarray info:
source = event_source(
datasets.get_dataset_path("gamma_test_large.simtel.gz"),
max_events=1,
)
event = next(iter(source))
# display the array
subarray = event.inst.subarray
ad = ArrayDisplay(subarray, tel_scale=3.0)
print("Now setting vectors")
plt.pause(1.0)
plt.tight_layout()
for phi in np.linspace(0, 360, 30) * u.deg:
r = np.cos(phi / 2)
ad.set_vector_rho_phi(r, phi)
plt.pause(0.01)
ad.set_vector_rho_phi(0, 0 * u.deg)
plt.pause(1.0)
print("Now setting values")
ad.telescopes.set_linewidth(0)
for ii in range(50):
vals = np.random.uniform(100.0, size=subarray.num_tels)
ad.values = vals
if step % 500 == 0:
print(step)
print(s.run(lrate))
print('sig,noise,len',s.run(sigma_2),s.run(noise_2),s.run(len_sc_2))
plt.clf()
plt.scatter(X,s.run(mean),color='blue')
#plt.scatter(s.run(tf.reshape(X_m_1,[M])),s.run(tf.reshape(opti_mu2,[M])),color='orange')
#plt.scatter(X_m_1,np.zeros((1,M)),color='purple')
plt.scatter(s.run(Xtr),s.run(X_mean2))
#plt.scatter(s.run(X_m),np.zeros(M),color='red')
plt.plot(X,s.run(mean),color='blue')
plt.scatter(X,Y,color='red')
#plt.plot(X,s.run(tf.add(var_terms,mean)),color='black')
#plt.plot(X,s.run(tf.sub(mean,var_terms)),color='black')
# plt.plot(s.run(Xtr),s.run(X_mean2+2*tf.sqrt(X_cov2+X_cov1)),color='green')
#plt.plot(s.run(Xtr),s.run(X_mean2-2*tf.sqrt(X_cov2+X_cov1)),color='green')
print('F2',s.run(F_v_2))
#plt.scatter(X,Y,color='green')
h.set_sess(s);GPLVM.printer()
plt.scatter(s.run(tf.reshape(X_m_2/10,[-1])),np.zeros(M),color='purple')
#plt.scatter(X_odd,Y_odd,color='black')
#plt.scatter(s.run(tf.reshape(X_m_1,[-1])),np.zeros(M),color='black')
plt.plot(s.run(Xtr),s.run(sample_all()),color='black')
plt.pause(2)
#print(s.run(tf.reduce_min(h.abs_diff(Z,Z)+np.eye(M,M))))
s.close()
plt.legend(loc='best')
# Calculate MTF.
# According to http://www.imatest.com/docs/validating_slanted_edge/, the MTF is
# the Fourier transform of the line spread function, which is the derivative of
# the averaged edge.
# We thus calculate the average edge ($source$Response) and take the derivative
# of this (with the `LSF` function, which uses `numpy.diff`).
# To the derivative we fit a gaussian function with the `gaussianfit` function
# which then acts as an input for the `MTF` function above.
plt.subplot(154)
plt.title('Line spread function')
plt.plot(LSF(ResponseERI), c=Colors[2], linestyle='dashed', label='ERI Data')
plt.plot(LSF(ResponseHamamatsu), c=Colors[3], linestyle='dashed', label='Hamamatsu Data')
plt.plot(gaussianfit(LSF(ResponseERI)), c=Colors[0], label='ERI fit')
plt.plot(gaussianfit(LSF(ResponseHamamatsu)), c=Colors[1], label='Hamamatsu fit')
plt.legend(loc='best')
plt.subplot(155)
plt.title('MTF')
plt.plot(normalize(MTF(gaussianfit(LSF(ResponseERI)))), c=Colors[0],
label='MTF ERI')
plt.plot(normalize(MTF(gaussianfit(LSF(ResponseHamamatsu)))), c=Colors[1],
label='MTF Hamamatsu')
plt.legend(loc='best')
# Save figure and concatenated results
plt.savefig(os.path.join(OutputPath, 'MTF%03dkV%03duA.png' % (VoltageERI[c], CurrentERI[c])))
plt.draw()
plt.pause(0.01)
plt.ioff()
plt.show()
import siena_cms_tools as cms
import matplotlib.pylab as plt
import sys
import time
start = time.time()
collisions = cms.get_collisions(sys.argv[1])
print "Time to read in %d collisions: %f seconds" % (len(collisions),time.time()-start)
print len(collisions)
#fig = plt.figure(figsize=(7,5),dpi=100)
#ax = fig.add_subplot(1,1,1)
#ax = fig.gca(projection='3d')
#plt.subplots_adjust(top=0.98,bottom=0.02,right=0.98,left=0.02)
fig = plt.figure()
#ax = fig.add_subplot(1,1,1)
ax = fig.gca(projection='3d')
for i in xrange(1000):
lines,fig,ax = cms.display_collision3D(collisions[i],fig=fig,ax=ax)
plt.pause(0.0001)
#plt.show(block=False)
image_extent = [0,num_cols*dx,0,num_rows*dx]
im = plt.imshow(zwR,cmap=plt.cm.RdBu,extent=image_extent)
plt.xlabel('Distance (m)', fontsize=12)
plt.ylabel('Distance (m)', fontsize=12)
# create contours
cset = plt.contour(zwR,extent=image_extent)
plt.clabel(cset,inline=True, fmt='%1.1f', fontsize=10)
# color bar on side
cb = plt.colorbar(im)
cb.set_label('Water Elevation (m)', fontsize=12)
# add title
plt.title('Water Elevation')
# Save plot
#plt.savefig('Water_Elevation')
# Display plot
plt.pause(0.1)
#else:
#pass
# FINALIZE
# get 2D array for contours
Hr = mg.node_vector_to_raster(H)
# Water height plot
image_extent = [0,num_cols*dx,0,num_rows*dx]
im = plt.imshow(Hr,cmap=plt.cm.RdBu,extent=image_extent)
plt.xlabel('Distance (m)', fontsize=12)
plt.ylabel('Distance (m)', fontsize=12)
