def histo_datalink(mode,
lambda_val,
attribute,
ci=95,
datalinks=range(0, NUM_DATA_LINK),
save=False):
stats = scalar_stats(scalar_parse(mode, lambda_val))
for u in datalinks:
attr = attribute + '-' + str(u)
bar = stats['mean'][attr]
error = np.array([
bar - stats['ci' + str(ci) + '_l'][attr],
stats['ci' + str(ci) + '_h'][attr] - bar
]).reshape(2, 1)
plt.bar('User ' + str(u),
bar,
yerr=error,
align='center',
alpha=0.95,
ecolor='k',
capsize=7)
# Show graphic
plt.title(attribute + ": " + MODE_DESCRIPTION[mode])
if save:
plt.savefig("histousers_" + attribute + "_" + mode + "_" + lambda_val +
".pdf",
bbox_inches="tight")
plt.clf()
else:
plt.show()
return
def correlationPGN(dataset):
allColsAsFeatures = dataset.columns.values[:-1]
X = dataset[allColsAsFeatures]
corr_matrix = X.corr()
# Generate a mask for the upper triangle
mask = np.zeros_like(corr_matrix, dtype=np.bool)
mask[np.triu_indices_from(mask)] = True
# Set up the matplotlib figure
f, ax = plt.subplots(figsize=(11, 9))
# Generate a custom diverging colormap
cmap = sns.diverging_palette(220, 10, as_cmap=True)
# Draw the heatmap with the mask and correct aspect ratio
sns_corrplot = sns.heatmap(corr_matrix,
mask=mask,
cmap=cmap,
vmax=1,
vmin=-1,
center=0,
square=True,
linewidths=.5,
cbar_kws={"shrink": .8})
sns_corrplot.get_figure()
path = WORKING_PATH + "/___corrrr.png"
plt.savefig(path)
plt.clf()
return path
def all_lorenz(mode,
lambda_val,
attribute,
iterations=range(0, NUM_ITERATIONS),
save=False):
data = scalar_parse(mode, lambda_val)
# Plot the mean lorenz
sel = data[data.name.str.startswith(attribute + '-')]
sel['user'] = sel.name.str.split('-', expand=True)[1].astype(int)
sorted_data = pd.DataFrame()
for r in iterations:
tmp = sel[sel.run == r]
sorted_data['run-' + str(r)] = np.sort(tmp.value.values)
plot_lorenz_curve(sorted_data['run-' + str(r)],
color='grey',
alpha=0.25)
# return sorted_data
plot_lorenz_curve(sorted_data.mean(axis=1))
plt.plot([0, 1], [0, 1], 'k', alpha=0.85)
plt.title(attribute + ": " + MODE_DESCRIPTION[mode] + ' - Mean Gini: ' +
str(gini(sorted_data.mean(axis=1))))
if save:
plt.savefig("lorenz_responseTime_" + mode + "_" + lambda_val + ".pdf")
plt.clf()
else:
plt.show()
return
def plotFeatImportance(pathOut,
imp,
oob,
oos,
method,
tag=0,
simNum=0,
**kargs):
# plot mean imp bars with std
mpl.figure(figsize=(10, imp.shape[0] / 5.))
imp = imp.sort_values('mean', ascending=True)
ax = imp['mean'].plot(kind='barh',
color='b',
alpha=0.25,
xerr=imp['std'],
error_kw={'ecolor': 'r'})
if method == 'MDI':
mpl.xlim([0, imp.sum(axis=1).max()])
mpl.axvline(1. / imp.shape[0], lw=1., color='r', ls='dotted')
ax.get_yaxis().set_visible(False)
for i, j in zip(ax.patches, imp.index):
ax.text(i.get_width() / 2,
i.get_y() + i.get_height() / 2,
j,
ha='center',
va='center',
color='k')
mpl.title('tag=' + tag + ' | simNUm=' + str(simNum) + ' | oob=' +
str(round(oob, 4)) + ' | oos=' + str(round(oos, 4)))
mpl.savefig(pathOut + 'featImportance_' + str(simNum) + '.png', dpi=100)
mpl.clf()
mpl.close()
return
def plot_stacked_barchart(self, dataframe, sort, title, xlable, ylable,
kurs):
x = []
tutor = []
y = []
for i in dataframe['tutor']:
if i not in tutor:
tutor.append(i)
y.append([])
for i, elem in enumerate(dataframe[sort]):
print(y, elem)
if elem in x:
y[tutor.index(dataframe['tutor'][i])][x.index(elem)] += 1
else:
x.append(elem)
for j, elem2 in enumerate(tutor):
y[j].append(0)
y[tutor.index(dataframe['tutor'][i])][x.index(elem)] += 1
for i, elem in enumerate(y):
plt.bar(range(len(elem)), elem, label=tutor[i])
plt.xlabel(xlable)
plt.ylabel(ylable)
plt.legend(loc="best")
plt.savefig('./PDFcreater/Plots/{}/{}.png'.format(kurs, title))
#plt.show()
# loescht den Plot fuer den Naechsten Plot
plt.clf()
plt.cla()
plt.close()
def histo(column):
mpg = pandas.read_csv("mpg.csv")
plt.clf()
if column in list(mpg.column):
plt.hist(column)
plt.title(column)
plt.savefig("static/histo.png")
else:
print("There is no such an attribute in the given data.")
return app.send_static_file("static/histo.png")
def plot_boring_barchart(self,dataframe,x,y,title,xlable,ylable, kurs):
plt.bar(x, y, color='blue')
#plt.title(title)
plt.xlabel(xlable)
plt.ylabel(ylable)
plt.savefig('./PDFcreater/Plots/{}/{}.png'.format(kurs,title))
#loescht den Plot fuer den Naechsten Plot
plt.clf()
plt.cla()
plt.close()
def save_fidelity(fname, fidelity):
steps = [
i * environment_configs.get_interval_width()
for i in range(len(fidelity))
]
plt.clf()
plt.plot(steps, fidelity)
plt.ylabel('fidelity')
plt.xlabel('step')
plt.savefig(fname + 'fidelity' + '_')
def plot_pie(self, nx_dataframe, topic, se_title, kurs):
gender = self.sort_column(nx_dataframe[topic])
labels = [gender[0][i] for i,elem in enumerate(gender[0])]
fracs = [gender[1][i] for i,elem in enumerate(gender[1])]
explode = [0.05 for i,elem in enumerate(gender[1])]
plt.pie(fracs, explode=explode, labels=labels, autopct='%.0f%%', shadow=True)
plt.savefig('./PDFcreater/Plots/{}/1{}.png'.format(kurs,se_title))
plt.clf()
plt.cla()
plt.close()
def save_controls(control_name, fname, action):
steps = [
i * environment_configs.get_interval_width()
for i in range(action.shape[0])
]
plt.clf()
plt.step(steps, action)
plt.ylabel(control_name)
plt.xlabel('t')
plt.savefig(fname + '_' + control_name)
def plot_training_val_accuracy(history, epochs):
plt.clf()
acc = history.history['acc']
val_acc = history.history['val_acc']
plt.plot(epochs, acc, 'g', label='Training acc')
plt.plot(epochs, val_acc, 'y', label='Validation acc')
plt.title('Training and validation accuracy')
plt.xlabel('Epochs')
plt.ylabel('Accuracy')
plt.legend()
plt.show()
def goplot(a, b):
amax = max(np.max(a), np.max(b))
amin = min(np.max(a), np.min(b))
plt.clf()
plt.plot([amin, amax], [amin, amax], 'k')
plt.plot(t[:, 0], t[:, 1], '.', color='green', markersize=5)
plt.plot(t[:, 0], t[:, 1], '.', color='orange', markersize=8)
plt.show()
def draw_heatmap(name, heatmap_data, x_labels, y_labels, cmap, center=0, path_to_dir=r""):
#maximum = 0
#minimum = 0
#for i in heatmap_data:
# for j in i:
# if j > maximum: maximum = j
# if j < minimum: minimum = j
#if center is None: plot = sns.heatmap(heatmap_data, xticklabels=x_labels, yticklabels=y_labels, robust=True, cmap=cmap, vmax=maximum, vmin=minimum)
plot = sns.heatmap(heatmap_data, xticklabels=x_labels, yticklabels=y_labels, center=0, vmin=-1, vmax=1, robust=True, cmap=cmap)
plt.savefig(Path(Path(path_to_dir) / Path("{}.png".format(name))))
plt.clf()
def plot(self, output):
plt.figure(figsize=output.fsize, dpi=output.dpi)
for ii in range(0, len(self.v)):
imsize = [self.t[0], self.t[-1], self.x[ii][-1], self.x[ii][0]]
lim = amax(absolute(self.v[ii])) / output.scale_sat
plt.imshow(self.v[ii], extent=imsize, vmin=-lim, vmax=lim, cmap=cm.gray, origin='upper', aspect='auto')
plt.title("%s-Velocity for Trace #%i" % (self.comp.upper(), ii))
plt.xlabel('Time (s)')
plt.ylabel('Offset (km)')
#plt.colorbar()
plt.savefig("Trace_%i_v%s.pdf" % (ii, self.comp))
plt.clf()
def plot_training_validation_loss(history):
plt.clf()
loss = history.history['loss']
val_loss = history.history['val_loss']
epochs = range(1, len(loss) + 1)
plt.plot(epochs, loss, 'g', label='Training loss')
plt.plot(epochs, val_loss, 'y', label='Validation loss')
plt.title('Training and validation loss')
plt.xlabel('Epochs')
plt.ylabel('Loss')
plt.legend()
plt.show()
def plotCentroidFitDiagnostic(img, hdr, ccdMod, ccdOut, res, prfObj):
"""Some diagnostic plots showing the performance of fitPrfCentroid()
Inputs:
-------------
img
(np 2d array) Image of star to be fit. Image is in the
format img[row, col]. img should not contain Nans
hdr
(Fits header object) header associated with the TPF file the
image was drawn from
ccdMod, ccdOut
(int) CCD module and output of image. Needed to
create the correct PRF model
prfObj
An object of the class prf.KeplerPrf()
Returns:
-------------
**None**
Output:
----------
A three panel subplot is created
"""
mp.figure(1)
mp.clf()
mp.subplot(131)
plotTpf.plotCadence(img, hdr)
mp.colorbar()
mp.title("Input Image")
mp.subplot(132)
c, r = res.x[0], res.x[1]
bbox = getBoundingBoxForImage(img, hdr)
model = prfObj.getPrfForBbox(ccdMod, ccdOut, c, r, bbox)
model *= res.x[2]
plotTpf.plotCadence(model, hdr)
mp.colorbar()
mp.title("Best fit model")
mp.subplot(133)
diff = img - model
plotTpf.plotCadence(diff, hdr)
mp.colorbar()
mp.title("Residuals")
print "Performance %.3f" % (np.max(np.abs(diff)) / np.max(img))
def plotCentroidFitDiagnostic(img, hdr, ccdMod, ccdOut, res, prfObj):
"""Some diagnostic plots showing the performance of fitPrfCentroid()
Inputs:
-------------
img
(np 2d array) Image of star to be fit. Image is in the
format img[row, col]. img should not contain Nans
hdr
(Fits header object) header associated with the TPF file the
image was drawn from
ccdMod, ccdOut
(int) CCD module and output of image. Needed to
create the correct PRF model
prfObj
An object of the class prf.KeplerPrf()
Returns:
-------------
**None**
Output:
----------
A three panel subplot is created
"""
mp.figure(1)
mp.clf()
mp.subplot(131)
plotTpf.plotCadence(img, hdr)
mp.colorbar()
mp.title("Input Image")
mp.subplot(132)
c,r = res.x[0], res.x[1]
bbox = getBoundingBoxForImage(img, hdr)
model = prfObj.getPrfForBbox(ccdMod, ccdOut, c, r, bbox)
model *= res.x[2]
plotTpf.plotCadence(model, hdr)
mp.colorbar()
mp.title("Best fit model")
mp.subplot(133)
diff = img-model
plotTpf.plotCadence(diff, hdr)
mp.colorbar()
mp.title("Residuals")
print "Performance %.3f" %(np.max(np.abs(diff))/np.max(img))
def main():
#DEFINITON OF THE PATHS TO THE FILES WITH THE CONTENT
path_to_train_accuracy = 'accuracy_train_data.csv'
path_to_train_loss = 'loss_train_data.csv'
path_to_validation_accuracy = 'accuracy_validation_data.csv'
path_to_validation_loss = '"loss_validation_data.csv"'
# CREATE LIST OF NUMBER OF EPOCHS COMPUTED
eval_indices = range(1, EPOCHS + 1)
#LOADS THE DATA FROM THE FILES
accuracy_train, loss_train, accuracy_validation, loss_validation = read_data(path_to_train_accuracy,path_to_train_loss,
path_to_validation_accuracy,path_to_validation_loss)
#SHOW THE INFORMATION FOR CONTROL OF QUALITY
print(eval_indices)
print("Accuracy Train: ",accuracy_train)
print("Loss Train: " ,loss_train)
print("Accuracy Validation: ", accuracy_validation)
print("Loss validation: ", loss_validation)
# DRAW THE ACCURACY GRAPH FOR VALIDATION AND TRAIN
plt.clf()
plt.subplot(211)
plt.plot(eval_indices, accuracy_train, 'k--', label='TREINO')
plt.plot(eval_indices, accuracy_validation, 'g-x', label='VALIDAÇÃO')
plt.legend(loc='upper right')
plt.xlabel('Épocas')
plt.ylabel('ACERTO')
plt.grid(which='major', axis='both')
# DRAW THE LOSS GRAPH FOR VALIDATION AND TRAIN
plt.subplot(212)
# plt.plot(eval_indices, train, 'g-x', label='Train Set Accuracy')
plt.plot(eval_indices, loss_train, 'r-x', label='TREINO')
# plt.plot(eval_indices, np.ones(len(eval_indices))/TOT_CLASSES, 'k--')
plt.plot(eval_indices, loss_validation, 'k--', label='VALIDAÇÃO')
plt.legend(loc="upper right")
plt.xlabel('Épocas')
plt.ylabel('ERRO')
plt.ylim(0, 1)
plt.grid(which='both', axis='y')
plt.subplots_adjust(left=0.2, wspace=0.2, hspace=0.3)
plt.show()
plt.pause(0.01)
#SAVES BOTH OF THE GRAPHICS IN ONE FILE NAMED "Learning.png"
plt.savefig('Learning.png')
def show_graph(lr_lists, epochs, steps, out_name='test'):
import matplotlib.pyplot as plt
plt.clf()
plt.rcParams['figure.figsize'] = [20, 5]
x = list(range(epochs * steps))
plt.plot(x, lr_lists, label="line L")
plt.plot()
plt.ylim(10e-5, 1)
plt.yscale("log")
plt.xlabel("iterations")
plt.ylabel("learning rate")
plt.title("Check Cosine Annealing Learing Rate with {}".format(out_name))
plt.legend()
plt.show()
def plot_feature_importances_cancer(scores):
names, val_scores = [name for name, _, _, _, _, _, _ in scores
], [score for _, score, _, _, _, _, _ in scores]
plt.rcParams["figure.figsize"] = [15, 9]
n_features = len(names)
plt.barh(range(n_features), val_scores, align='center')
plt.yticks(np.arange(n_features), names)
plt.xlabel("Accuracy")
plt.ylabel("Model")
path = WORKING_PATH + "/___comparison_cancer_model.png"
plt.savefig(path)
plt.clf()
return path
def show_acc(history):
""" 绘制精度曲线 """
plt.clf()
history_dict = history.history
acc = history_dict['binary_accuracy']
val_acc = history_dict['val_binary_accuracy']
epochs = range(1, len(val_acc) + 1)
plt.plot(epochs, acc, 'bo', label='Training acc')
plt.plot(epochs, val_acc, 'b', label='Validation acc')
plt.xlabel('Epochs')
plt.ylabel('Acc')
plt.legend()
plt.show()
def graph(x,y,xLabel,yLabel,title,figname):
plt.clf()
plt.hist(x,color="c",edgecolor="k",alpha=0.5)
plt.axvline(np.array(x).mean(),color="k",linestyle="dashed",linewidth=3,label="average")
plt.xlabel(xLabel)
plt.ylabel(yLabel)
plt.title(title)
yAxis = np.arange(0,10,1)
acRes = [y]
z = np.array(acRes*10)
plt.plot(z,yAxis,label="model accuracy")
p_value = ttest_ind(x,[y])[1]
plt.plot([],[],label=f"p-value: {np.round(p_value,4)}",color="w")
plt.legend()
plt.savefig(figname)
def make_mean_carrier(dm):
#Create the statistics on a data set sorted also by sector.
mean_count = dm['Carrier'].mean()
if mean_count.dtype == float and np.isnan(mean_count):
mean_count = 0
std_count = dm['Carrier'].std()
if std_count.dtype == float and np.isnan(std_count):
std_count = 0
max_count = dm['Carrier'].max()
min_count = dm['Carrier'].min()
top_count_sector = mean_count + std_count
bad_sectors = dm.loc[dm['Carrier']>top_count_sector]
sector_count = float(len(dm))
try:
bad_sectors = float(bad_sectors)
except:
bad_sectors =0
#Dropped Call Failure reason
#Call Final Class qualifier
print '++++++++++++++++++++\n','Total Sector Count = %4d'%(sector_count)
print '\nNumber of Top Offending Sectors = %4d '%bad_sectors
print '%4.2f percent'%((100)*(bad_sectors/sector_count))
# Find the cutoff statistics
print '\nMean =%4d'%(mean_count), 'Standard deviation ',std_count, 'Mean + 1 sigma = %4d'%(top_count_sector)
if False:
fig =plt.figure()
plt.clf()
bin_tick =np.arange(top_count_sector,max_count,50, dtype=int)
#dm2.hist()
#dm2.plot(x=['ECP','Cell'], y ='Carrier',kind='line')
#dm.plot(x=['ECP','Cell'], y='Carrier' )
#plt.subplots(2,2)
ax1 = fig.add_subplot(2,2,1)
_ = ax1.plot(dm['Carrier'].values, drawstyle='steps-post', label ='steps-post')
ax2 = fig.add_subplot(2,2,3)
bin_tick = np.arange(min_count,max_count,50, dtype=int)
_ = ax2.hist(dm['Carrier'],bins=bin_tick)
ax3 = fig.add_subplot(2,2,4)
norm_cdf(68, std_count, min_count, max_count)
return top_count_sector
def plot_facet(name: str, X, cluster_predictions):
temp = X.copy().reset_index(drop=True)
temp[f"{name}_cluster"] = cluster_predictions
temp = temp.melt(id_vars=f"{name}_cluster")
means = temp.groupby([f"{name}_cluster", "variable"]).mean().reset_index()
g = sns.FacetGrid(means,
col="variable",
hue=f"{name}_cluster",
col_wrap=5,
height=2,
sharey=False)
g = g.map(plt.bar, f"{name}_cluster", "value").set_titles("{col_name}")
g.savefig(f"outputs/{name}_facetgrid.png")
plt.clf()
return
def plot(self, output):
plt.figure(figsize=output.fsize, dpi=output.dpi)
for ii in range(0, len(self.v)):
imsize = [self.t[0], self.t[-1], self.x[ii][-1], self.x[ii][0]]
lim = amax(absolute(self.v[ii])) / output.scale_sat
plt.imshow(self.v[ii],
extent=imsize,
vmin=-lim,
vmax=lim,
cmap=cm.gray,
origin='upper',
aspect='auto')
plt.title("%s-Velocity for Trace #%i" % (self.comp.upper(), ii))
plt.xlabel('Time (s)')
plt.ylabel('Offset (km)')
#plt.colorbar()
plt.savefig("Trace_%i_v%s.pdf" % (ii, self.comp))
plt.clf()
def plot_rewards(self):
plt.figure(1)
plt.clf()
plt.title('Training...')
plt.xlabel('Time')
plt.ylabel('Reward')
cumulative = []
for i in range(len(self.rewards[self.n_training])):
if i == 0:
cumulative.append(self.rewards[self.n_training][i])
else:
cumulative.append(self.rewards[self.n_training][i] +
cumulative[-1])
x = np.linspace(2.0,
len(self.rewards[self.n_training]),
num=len(self.rewards[self.n_training]))
plt.plot(x, cumulative)
plt.show()
plt.pause(0.001) # pause a bit so that plots are updated
def getHistogram2PGN(df):
y = df.iloc[:, -1]
c = Counter(y)
numDiffClasses = len(y.unique())
target_names = y.unique()
colors = colors = ['lightcoral', 'gold', 'yellowgreen',
'cyan'][:numDiffClasses]
plt.rcParams["figure.figsize"] = [10, 5]
plt.pie([c[i] / len(y) * 100.0 for i in c],
labels=target_names,
colors=colors,
autopct='%1.1f%%',
shadow=True,
startangle=90)
plt.axis('equal')
plt.title(df.columns.values[-1])
path = WORKING_PATH + "/___circle_labels.png"
plt.savefig(path)
plt.clf()
return path
def plotLoss(self):
""" """
loss = self.loss_log
y0, y1 = [], []
for row in loss:
y0.append(float(row[0]))
y1.append(float(row[1]))
if len(y0) == 0:
return
window_size = 100
window = np.ones(int(window_size))/float(window_size)
y_av0 = np.convolve(y0, window, 'same')
y_av1 = np.convolve(y1, window, 'same')
arr = np.array(y_av0)
plt.clf() # Clear.
plt.title("loss")
plt.plot(y_av0[:-50])
#plt.plot(y_av1[:-50])
plt.ylabel('Smoothed Loss')
plt.show()
def checkCircle(data):
import matplotlib as plot
# Transform data by rotation so that all data appears in the firt and
# second quadrants (so that positive square root solution is valid)
rot = np.arctan2(data[-1, 1] - data[0, 1], data[-1, 0] - data[0, 0])
R = getRotMat(rot + np.pi)
trans = np.matmul(data, R)
trans[:, 1] *= -1
# Display transformed data used for the actual fit
plot.figure(2)
plot.clf()
plot.plot(trans[:, 0], trans[:, 1], '*')
plot.axis("equal")
thresh = 0.01
E = 10
count = 0
# Perform gradient descent up to four times with slightly different guesses
# until a low E solution is found.
while E > thresh and count < 4:
count += 1
xBar = np.mean(trans[:, 0]) * (0.95 + np.random.rand(1) * 0.1)[0]
yBar = np.min(trans[:, 1]) * (0.95 + np.random.rand(1) * 0.1)[0]
rBar = np.abs((trans[0, 0] - trans[-1, 0]) /
2) * (0.95 + np.random.rand(1) * 0.1)[0]
x0, y0, r, E = descent(trans, xBar, yBar, rBar)
print(E)
# More plotting of transformed data
circle = plot.Circle((x0, y0), r, color='r', fill=False)
plot.gca().add_artist(circle)
# Rotate back to original coordinates
y0 *= -1
center = np.matmul([x0, y0], np.linalg.inv(R))
if E < thresh:
return center, r
else:
return None, None
def exampleDiffImgCentroiding():
k2id = 206103150
campaign = 3
ar = mastio.K2Archive()
fits, hdr = ar.getLongTpf(k2id, campaign, header=True)
hdr0 = ar.getLongTpf(k2id, campaign, ext=0)
cube = tpf.getTargetPixelArrayFromFits(fits, hdr)
idx = np.isfinite(cube)
cube[~idx] = 0 #Remove Nans
flags = fits['QUALITY']
ccdMod = hdr0['module']
ccdOut = hdr0['output']
#Compute roll phase
llc = ar.getLongCadence(k2id, campaign)
time= llc['TIME']
cent1 = llc['MOM_CENTR1']
cent2 = llc['MOM_CENTR2']
centColRow = np.vstack((cent1, cent2)).transpose()
rot = arclen.computeArcLength(centColRow, flags>0)
rollPhase = rot[:,0]
rollPhase[flags>0] = -9999 #A bad value
prfObj = prf.KeplerPrf("/home/fergal/data/keplerprf")
bbox = getBoundingBoxForImage(cube[0], hdr)
period = 4.1591409
epoch = fits['time'][491]
dur = 3.0
out, log = measureDiffOffset(period, epoch, dur, time, prfObj, \
ccdMod, ccdOut, cube, bbox, rollPhase, flags)
idx = out[:,1] > 0
mp.clf()
mp.plot(out[:,3]-out[:,1], out[:,4]- out[:,2], 'ro')
return out
def ecdf_sca(data,
attribute,
aggregate=False,
datalinks=range(0, NUM_DATA_LINK),
save=False):
if aggregate:
selected_ds = data[data.name.isin([
attribute + '-' + str(i) for i in datalinks
])].groupby('run').mean()
else:
selected_ds = data[data.name == attribute]
plot_ecdf(selected_ds.value.to_numpy())
plt.title("ECDF for " + attribute +
(" (aggregated mean)" if aggregate else ""))
if save:
plt.savefig("ecdf_" + attribute + ".pdf", bbox_inches="tight")
plt.clf()
else:
plt.show()
return
def exampleDiffImgCentroiding():
k2id = 206103150
campaign = 3
ar = mastio.K2Archive()
fits, hdr = ar.getLongTpf(k2id, campaign, header=True)
hdr0 = ar.getLongTpf(k2id, campaign, ext=0)
cube = tpf.getTargetPixelArrayFromFits(fits, hdr)
idx = np.isfinite(cube)
cube[~idx] = 0 #Remove Nans
flags = fits['QUALITY']
ccdMod = hdr0['module']
ccdOut = hdr0['output']
#Compute roll phase
llc = ar.getLongCadence(k2id, campaign)
time = llc['TIME']
cent1 = llc['MOM_CENTR1']
cent2 = llc['MOM_CENTR2']
centColRow = np.vstack((cent1, cent2)).transpose()
rot = arclen.computeArcLength(centColRow, flags > 0)
rollPhase = rot[:, 0]
rollPhase[flags > 0] = -9999 #A bad value
prfObj = prf.KeplerPrf("/home/fergal/data/keplerprf")
bbox = getBoundingBoxForImage(cube[0], hdr)
period = 4.1591409
epoch = fits['time'][491]
dur = 3.0
out, log = measureDiffOffset(period, epoch, dur, time, prfObj, \
ccdMod, ccdOut, cube, bbox, rollPhase, flags)
idx = out[:, 1] > 0
mp.clf()
mp.plot(out[:, 3] - out[:, 1], out[:, 4] - out[:, 2], 'ro')
return out
def plot_scores(scores):
plt.figure(figsize=(15, 6))
names, val_scores = [name for name, _, _, _, _, _, _ in scores
], [score for _, score, _, _, _, _, _ in scores]
ax = sns.barplot(x=names, y=val_scores)
for p, score in zip(ax.patches, val_scores):
height = p.get_height()
ax.text(p.get_x() + p.get_width() / 2.,
height + 0.005,
'{:1.3f}'.format(score),
ha="center",
fontsize=14)
plt.xlabel('method', fontsize=18)
plt.ylabel('Mean Val. Accuracy', fontsize=18)
plt.xticks(rotation=90, fontsize=16)
plt.yticks(fontsize=16)
plt.ylim(0.6, 1)
path = WORKING_PATH + "/___comparison.png"
plt.savefig(path)
plt.clf()
return path
def plotdatatree(treeID, scale1, mass1):
plot_title="Mass Accretion History Tree " + str(treeID) #Can code the number in with treemax
x_axis="scale time"
y_axis="total mass"
figure_name=os.path.expanduser('~/figureTree' + str(treeID))
#Choose which type of plot you would like: Commented out.
plt.plot(scale1, mass1, linestyle="-", marker="o")
#plt.scatter(scale1, mass1, label="first tree")
plt.title(plot_title)
plt.xlabel(x_axis)
plt.ylabel(y_axis)
#plt.yscale("log")
plt.savefig(figure_name)
#In order to Plot only a single tree on a plot must clear lists before loop.
#Comment out to over plot curves.
plt.clf()
clearmass = []
clearscale = []
return clearmass, clearscale
def train_dcgan_labeled(gen, dis, epoch0=0):
print('CHAINER begin training'); sys.stdout.flush()
o_gen = optimizers.Adam(alpha=0.0002, beta1=0.5)
o_dis = optimizers.Adam(alpha=0.0002, beta1=0.5)
print('CHAINER begin gen'); sys.stdout.flush()
o_gen.setup(gen)
o_dis.setup(dis)
print('CHAINER begin add'); sys.stdout.flush()
o_gen.add_hook(chainer.optimizer.WeightDecay(0.00001))
o_dis.add_hook(chainer.optimizer.WeightDecay(0.00001))
print('CHAINER begin zvis'); sys.stdout.flush()
# zvis = (xp.random.uniform(-1, 1, (100, nz), dtype=np.float32))
print('CHAINER begin for'); sys.stdout.flush()
for epoch in range(epoch0,n_epoch):
print("epoch:",epoch)
sys.stdout.flush()
perm = np.random.permutation(n_train)
sum_l_dis = np.float32(0)
sum_l_gen = np.float32(0)
for i in range(0, n_train, batchsize):
# discriminator
# 0: from dataset
# 1: from noise
#print "load image start ", i
x2 = np.zeros((batchsize, 3, 96, 96), dtype=np.float32)
for j in range(batchsize):
#try:
rnd = np.random.randint(len(dataset))
rnd2 = np.random.randint(2)
img = np.asarray(Image.open(io.BytesIO(dataset[rnd])).convert('RGB')).astype(np.float32).transpose(2, 0, 1)
x2[j,:,:,:] = (img[:,0:96,0:96]-128.0)/128.0
#except:
# print('read image error occured', fs[rnd])
#print "load image done"
# train generator
z = Variable(xp.random.uniform(-1, 1, (batchsize, nz), dtype=np.float32))
x = gen(z)
yl = dis(x)
L_gen = F.softmax_cross_entropy(yl, Variable(xp.zeros(batchsize, dtype=np.int32)))
L_dis = F.softmax_cross_entropy(yl, Variable(xp.ones(batchsize, dtype=np.int32)))
# train discriminator
x2 = Variable(cuda.to_gpu(x2))
yl2 = dis(x2)
L_dis += F.softmax_cross_entropy(yl2, Variable(xp.zeros(batchsize, dtype=np.int32)))
#print "forward done"
o_gen.zero_grads()
L_gen.backward()
o_gen.update()
o_dis.zero_grads()
L_dis.backward()
o_dis.update()
sum_l_gen += L_gen.data.get()
sum_l_dis += L_dis.data.get()
#print "backward done"
if i%image_save_interval==0:
pylab.rcParams['figure.figsize'] = (16.0,16.0)
pylab.clf()
vissize = 100
z = zvis
z[50:,:] = (xp.random.uniform(-1, 1, (50, nz), dtype=np.float32))
z = Variable(z)
x = gen(z, test=True)
x = x.data.get()
for i_ in range(100):
tmp = ((np.vectorize(clip_img)(x[i_,:,:,:])+1)/2).transpose(1,2,0)
pylab.subplot(10,10,i_+1)
pylab.imshow(tmp)
pylab.axis('off')
pylab.savefig('%s/vis_%d_%d.png'%(out_image_dir, epoch,i))
serializers.save_hdf5("%s/dcgan_model_dis_%d.h5"%(out_model_dir, epoch),dis)
serializers.save_hdf5("%s/dcgan_model_gen_%d.h5"%(out_model_dir, epoch),gen)
serializers.save_hdf5("%s/dcgan_state_dis_%d.h5"%(out_model_dir, epoch),o_dis)
serializers.save_hdf5("%s/dcgan_state_gen_%d.h5"%(out_model_dir, epoch),o_gen)
print('epoch end', epoch, sum_l_gen/n_train, sum_l_dis/n_train)
#find CIs, using ssm
ci_upper=np.zeros( (1,100))
ci_lower=np.zeros( (1,100))
m_boot=np.zeros( (1,100))
for i in range(100):
ci_upper[0,i]=ssm.scoreatpercentile(bootdata[i,:],97.5)
ci_lower[0,i]=ssm.scoreatpercentile(bootdata[i,:],02.5)
m_boot[0,i]=np.mean(bootdata[i,:])
print "PLOTTING"
# plot -------------------------------------------
# thank you tomas http://www.staff.ncl.ac.uk/tom.holderness/software/pythonlinearfit
plt.clf()
# plot sample data
plot(plot_timespread,'ro',label='Sample observations')
# plot line of best fit
plot(m_boot[(0,)],'b-',label='bootstrap_mean')
# plot confidence limits
plot(ci_lower[(0,)],'b--',label='confidence limits (95%)')
plot(ci_upper[(0,)],'b--')
# configure legend
legend(loc=4) #lower left http://matplotlib.org/users/legend_guide.html
leg = gca().get_legend()
ltext = leg.get_texts()
diabetes_y_train = diabetes.target[:-20]
diabetes_y_test = diabetes.target[-20:]
regr = linear_model.LinearRegression()
regr.fit(diabetes_X_train, diabetes_y_train)
# >> LinearRegression(copy_X=True, fit_intercept=True, normalize=False)
print regr.coef_
# The mean square error
np.mean((regr.predict(diabetes_X_test)-diabetes_y_test)**2)
# Explained variance score: 1 is perfect prediction
# and 0 means that there is no linear relationship
# between X and Y.
regr.score(diabetes_X_test, diabetes_y_test)
# Plot results
pl.clf() # Clear plots
pl.plot(diabetes_X_test, regr.fit(diabetes_X_train, diabetes_y_train));
pl.title('Linear regression of sample diabetes data\n'
'Centroids are marked with white cross')
pl.xlim(x_min, x_max)
pl.ylim(y_min, y_max)
pl.xticks(())
pl.yticks(())
pl.show()
