def plot(x_param, x_label, y_param, y_label, linear=False):
global currentFig
spear = sp.spearmanr(x_param, y_param, nan_policy='omit')
print("Figure %2.1d %13s vs %-13s Spearman: %8.3g pvalue: %8.2g" %
(currentFig, y_label, x_label, spear[0], spear[1]))
fig = plt.figure(currentFig) # the current figure
currentFig += 1
plt.clf() # clear the figure before each run
ax = fig.add_subplot(111) # set axes, figure location
if linear == True:
ax.plot(x_param, y_param, 'o')
else:
ax.loglog(x_param, y_param, 'o')
ax.set_xlabel("%s" % LABELS[x_label], fontsize=15)
ax.set_ylabel("%s" % LABELS[y_label], fontsize=15)
plt.tight_layout()
plt.show()
return
def main():
parser = argparse.ArgumentParser()
parser.add_argument('-p',
'--dimension',
type=int,
default=DIMENSION_DEFAULT)
parser.add_argument('-N',
'--num-samples',
type=int,
default=[60, 70, 80, 100, 150, 250, 500, 850],
nargs='+')
parser.add_argument('-M',
'--average-iterations',
type=int,
default=AVERAGE_ITERATIONS_DEFAULT)
parser.add_argument('-b',
'--beta',
type=float,
default=[1.0, 0.5, 0.2],
nargs='+')
args = parser.parse_args()
p = args.dimension
Ns = args.num_samples
M = args.average_iterations
T = 200
TYLER_MAX_ITERS = 1000
TYLER_NEWTON_STEPS = 750
GAUSSIAN_NEWTON_STEPS = 750
for beta in args.beta:
dataset = SyntheticDataset(p=p, Ns=Ns, M=M, beta=beta)
metric = JointEstimationDistanceErrorMetric(T=T, dataset=dataset)
estimator_objects = [
#estimators.joint.general_loss.MMNewtonJointEstimator(
# loss=losses.tyler(dataset.get_dimension()), tolerance=1e-6, max_iters=TYLER_MAX_ITERS, newton_num_steps=TYLER_NEWTON_STEPS,
# newton_tol=1e-6, name='Tyler'),
# estimators.joint.gauss_loss.NewtonJointEstimator(
# newton_num_steps=GAUSSIAN_NEWTON_STEPS, newton_tol=1e-6, name='GMRF Newton'),
# estimators.joint.general_loss.MMJointEstimator(
# estimators.joint.gauss_loss.InvestJointEstimator(), loss=losses.tyler(dataset.get_dimension()),
# tolerance=1e-6, max_iters=TYLER_MAX_ITERS, name='Tyler'),
#estimators.joint.gauss_loss.InvestJointEstimator(name='GMRF'),
estimators.joint.general_loss.MMNewtonJointEstimator(
loss=losses.generalized_gaussian(beta, 1),
tolerance=1e-6,
max_iters=TYLER_MAX_ITERS,
newton_num_steps=TYLER_NEWTON_STEPS,
newton_tol=1e-6,
name='GG'),
estimators.joint.gauss_loss.SampleCovarianceJointEstimator(
name='Sample covariance')
]
plots.plot_variables_vs_N(estimator_objects, Ns, M, metric, show=False)
plots.show()
def plots(self):
# traded_volume_over_time(self.num_steps, self.agent_measurements)
plots.soc_over_time(self.num_steps, self.soc_list_over_time)
plots.households_deficit_overflow(self.num_steps, self.deficit_over_time, self.overflow_over_time)
plots.clearing_over_utility_price(
self.num_steps, self.utility_price, self.clearing_price_min_avg_max, self.clearing_quantity)
plots.clearing_quantity(self.num_steps, self.clearing_quantity)
plots.clearing_quantity_over_demand(self.num_steps, self.clearing_quantity, self.household_demand)
# If an electrolyzer is present, plot its behaviour.
if 'Electrolyzer' in self.model.agents:
plots.electrolyzer(self.num_steps, self.model.agents['Electrolyzer'])
plots.show()
def main():
datas = map(read_data_json, sys.argv[1:])
# draw_all(datas, 'reactions', [
# 'forward methyl to high bridge',
# 'reverse methyl to high bridge',
# ])
draw_all(datas, 'species', [
# re.compile(r'^(?!dimer).*$')
'high bridge',
'methyl on dimer'
])
show()
def runTest(xTrain, yTrain, xTest, yTest, arguments, label="NA", index=None):
(trainX, trainY) = loadData(xTrain, yTrain)
(testX, testY) = loadData(xTest, yTest)
# required scaling for rbm
trainX = scale(trainX)
testX = scale(testX)
# save the test concentration series, testY becomes class labels
testC = testY
if (arguments["multiClass"] == 1):
trainY = convertToClasses(trainY)
testY = convertToClasses(testY)
else:
trainY = np.transpose(trainY)
testY = np.transpose(testY)
logitPred = doLogisticRegression(trainX, trainY, testX, testY, arguments["optimize"], arguments["pickle"])
rbmPred = doRBM(trainX, trainY, testX, testY, arguments["optimize"], arguments["pickle"])
# write the results of the training / optimization phase
if (arguments["optimize"] == "new" or arguments["optimize"] == "gs"):
if (not os.path.isfile(arguments["label"] + 'train_result.csv')):
mode = 'w'
writeH = True
else:
mode = 'a'
writeH = False
with open(arguments["label"] + 'train_result.csv', mode) as csvfile:
header = ['index', 'model_fit', 'rbm_accuracy', 'logit_accuracy']
label = index if not (index==None) else np.mean(testC[:,0])
writer = csv.DictWriter(csvfile, fieldnames=header)
if (writeH): writer.writeheader()
writer.writerow({'index' : label,
'model_fit' : arguments["optimize"],
'rbm_accuracy' : accuracy_score(testY, rbmPred),
'logit_accuracy' : accuracy_score(testY, logitPred)
})
print [label, accuracy_score(testY, rbmPred), accuracy_score(testY, logitPred)]
if (arguments["verbose"] == 1):
print "LOGISTIC REGRESSION PERFORMANCE"
print classification_report(testY, logitPred)
print("Accuracy Score: %s\n" % accuracy_score(testY, logitPred))
print "RBM PERFORMANCE"
print classification_report(testY, rbmPred)
print("Accuracy Score: %s\n" % accuracy_score(testY, rbmPred))
if (arguments["visualize"] == 1):
plot.accuracy(testY, logitPred, "Logistic Regression", c=testC)
plot.accuracy(testY, rbmPred, "RBM", c=testC)
plot.show()
if (arguments["predOut"] == 1):
np.savetxt("logitPred.csv", logitPred, delimiter=",")
np.savetxt("rbmPred.csv", rbmPred, delimiter=",")
if (index==None):
label = np.mean(testC[:,0])
else:
label = index
predictions = [(index, 'logistic', logitPred), (index, 'rbm', rbmPred)]
rets = []
for l,c,p in predictions:
ret = {'label': l,
'clf': c,
'accuracy_score': accuracy_score(testY, p),
'precision_score': precision_score(testY, p),
'recall_score': recall_score(testY, p),
'f1_score': f1_score(testY, p)}
rets.append(ret)
return rets
import message import mixer import channel import receiver from plots import plot_before_after, plot_before_after_amqam, show # message orig1 = message.sin() orig2 = message.square() m1 = message.sin() m2 = message.square() # mixer m = mixer.AMQAM(m1, m2) # channel m = channel.attenuation(m) m = channel.fading(m) m = channel.gauss_noise(m) # receiver m = receiver.unattenuate(m) m1, m2 = receiver.demuxAMQAM(m) m1 = receiver.lpf(m1) m2 = receiver.lpf(m2) plot_before_after_amqam(orig1, orig2, m1, m2) show()
end = args["end"]
start = args["start"]
data = np.absolute(data)
ctx_pred = data[start:end, 1]
f_pred = data[start:end, 0]
target = data[start:end, 2]
ctx_accuracy = (target == ctx_pred)
f_accuracy = (target == f_pred)
if (not args["concentration"] == "none"):
c = np.genfromtxt(args["concentration"], delimiter=",", dtype="float32")
else:
c = np.zeros((0, 0))
print("PREDICTION FROM CORTEX")
print classification_report(target, ctx_pred)
print("Accuracy Score: %s\n" % accuracy_score(target, ctx_pred))
print("PREDICTION FROM FIBERS")
print classification_report(target, f_pred)
print("Accuracy Score: %s\n" % accuracy_score(target, f_pred))
if (args["visualize"] == 1):
plot.accuracy(target, ctx_pred, label="Cortex", c=c)
plot.accuracy(target, f_pred, label="Fibers", c=c)
plot.show()
# prediction for the generation of the confusion matrix
y_pred_dumm = pd.DataFrame(
model.predict(X_test_reshaped, batch_size=100),
columns=y_test_dumm.columns,
)
y_pred = y_pred_dumm.idxmax(axis="columns").astype("category")
precision, recall, fscore, support = precision_recall_fscore_support(
y_test, y_pred)
val = (
y_test.value_counts().rename("support").to_frame().reset_index().merge(
pd.DataFrame({
"precision": precision,
"recall": recall,
"fscore": fscore,
"support": support,
}),
on="support",
).set_index("index"))
print(val)
# confusion matrix
plots.confusion_matrix_tf(y_test, y_pred)
plots.show()
# print the architecture
# from tensorflow.keras.utils import plot_model
# plot_model(model, to_file="model.png", show_shapes=True, show_layer_names=True)
start = args["start"]
data = np.absolute(data)
ctx_pred = data[start:end,1]
f_pred = data[start:end,0]
target = data[start:end,2]
ctx_accuracy = (target == ctx_pred)
f_accuracy = (target == f_pred)
if (not args["concentration"] == "none"):
c = np.genfromtxt(args["concentration"], delimiter=",", dtype="float32")
else:
c = np.zeros((0,0))
print("PREDICTION FROM CORTEX")
print classification_report(target, ctx_pred)
print("Accuracy Score: %s\n" % accuracy_score(target, ctx_pred))
print("PREDICTION FROM FIBERS")
print classification_report(target, f_pred)
print("Accuracy Score: %s\n" % accuracy_score(target, f_pred))
if (args["visualize"] == 1):
plot.accuracy(target, ctx_pred, label="Cortex", c=c)
plot.accuracy(target, f_pred, label="Fibers", c=c)
plot.show()
bgcounts = np.recfromtxt(args.bgfile, comments="#", delimiter="\t", dtype=int, names=True)
counts = np.recfromtxt(args.file, comments="#", delimiter="\t", names=True, dtype=int)
dp = DataPrep(counts, bgcounts, crop=args.crop, fill=args.fill, normalize=False, down_sampling=args.dsfac)
dp.gap_stats()
if args.lomb_scargle:
# calculate the LombScargle Periodigram in the
# frequency range of interessest
freq_ls = np.linspace(5.0e-7, 5.0e-4, 500)
ls = LombScargle(np.float_(dp.position), np.float_(dp.counts), freq_ls)
f, pgram = ls()
ax = plots.samples(f, pgram, window_title="Lomb-Scargle Periodigram")
ax.semilogy()
if args.welch:
plots.samples(dp.position, dp.counts, window_title="Raw data")
# xlim = (0.0, 5e-5)
fs = 1.0 / np.diff(dp.position).min()
plots.psd(
dp.counts,
xlim=args.frange,
Fs=fs,
NFFT=args.nfft,
window=mlab.window_none,
window_title="Welch Periodigram",
)
plots.samples(dp.position, dp.counts, window_title="Raw data")
plots.show()