From 531badffc81970f0b8bf48631594062105f0aa21 Mon Sep 17 00:00:00 2001
From: Alejandro Moreo <alejandro.moreo@isti.cnr.it>
Date: Wed, 25 Sep 2024 17:00:22 +0200
Subject: [PATCH] change in minimize -> minimize_scalar in TKDE2
---
KDEy/kdey_devel.py | 26 +++++++++++++++-----------
KDEy/quantification_evaluation.py | 4 ++--
2 files changed, 17 insertions(+), 13 deletions(-)
diff --git a/KDEy/kdey_devel.py b/KDEy/kdey_devel.py
index fa3713b..0506fde 100644
--- a/KDEy/kdey_devel.py
+++ b/KDEy/kdey_devel.py
@@ -68,11 +68,11 @@ class KDEyMLauto(KDEyML):
current_prevalence, current_bandwidth = self.optim_minimize_both_fine(current_bandwidth, current_prevalence, tr_posteriors, tr_y, te_posteriors, classes)
# check converngece
- prev_convergence = all(np.isclose(previous_prevalence, current_prevalence, atol=0.0001))
+ prev_convergence = all(np.isclose(previous_prevalence, current_prevalence, atol=0.01))
if isinstance(current_bandwidth, float):
- band_convergence = np.isclose(previous_bandwidth, current_bandwidth, atol=0.0001)
+ band_convergence = np.isclose(previous_bandwidth, current_bandwidth, atol=0.001)
else:
- band_convergence = all(np.isclose(previous_bandwidth, current_bandwidth, atol=0.0001))
+ band_convergence = all(np.isclose(previous_bandwidth, current_bandwidth, atol=0.001))
convergence = band_convergence and prev_convergence
@@ -94,17 +94,21 @@ class KDEyMLauto(KDEyML):
def optim_minimize_bandwidth(self, current_bandwidth, current_prev, tr_posteriors, tr_y, te_posteriors, classes):
def neg_loglikelihood_bandwidth(bandwidth):
- mix_densities = self.get_mixture_components(tr_posteriors, tr_y, classes, bandwidth[0])
+ # bandwidth = bandwidth[0]
+ mix_densities = self.get_mixture_components(tr_posteriors, tr_y, classes, bandwidth)
test_densities = [self.pdf(kde_i, te_posteriors) for kde_i in mix_densities]
test_mixture_likelihood = sum(prev_i * dens_i for prev_i, dens_i in zip(current_prev, test_densities))
test_loglikelihood = np.log(test_mixture_likelihood + epsilon)
- return -np.sum(test_loglikelihood)
+ nll = -np.sum(test_loglikelihood)
+ # print(f'\t{bandwidth=:.10f}\t{nll=:.10f}')
+ return nll
- bounds = [(0.00001, 1)]
- r = optimize.minimize(neg_loglikelihood_bandwidth, x0=[current_bandwidth], method='SLSQP', bounds=bounds)
- print(f'iterations-bandwidth={r.nit}')
- assert r.success, 'Process did not converge!'
- return r.x[0]
+ # bounds = [(0.00001, 0.2)]
+ # r = optimize.minimize(neg_loglikelihood_bandwidth, x0=[current_bandwidth], method='SLSQP', bounds=bounds)
+ r = optimize.minimize_scalar(neg_loglikelihood_bandwidth, bounds=(0.00001, 0.2))
+ # print(f'iterations-bandwidth={r.nit}')
+ assert r.success, f'Process did not converge! {r.message}'
+ return r.x
def optim_minimize_both(self, current_bandwidth, current_prev, tr_posteriors, tr_y, te_posteriors, classes):
n_classes = len(current_prev)
@@ -117,7 +121,7 @@ class KDEyMLauto(KDEyML):
test_loglikelihood = np.log(test_mixture_likelihood + epsilon)
return -np.sum(test_loglikelihood)
- bounds = [(0, 1) for _ in range(n_classes)] + [(0.00001, 1)]
+ bounds = [(0, 1) for _ in range(n_classes)] + [(0.00001, 0.2)]
constraints = ({'type': 'eq', 'fun': lambda x: 1 - sum(x[:n_classes])})
prevalence_bandwidth = np.append(current_prev, current_bandwidth)
r = optimize.minimize(neg_loglikelihood_bandwidth, x0=prevalence_bandwidth, method='SLSQP', bounds=bounds, constraints=constraints)
diff --git a/KDEy/quantification_evaluation.py b/KDEy/quantification_evaluation.py
index 9cc8a8e..bfe8f58 100644
--- a/KDEy/quantification_evaluation.py
+++ b/KDEy/quantification_evaluation.py
@@ -35,7 +35,7 @@ METHODS = [
('PACC', PACC(newLR()), wrap_hyper(logreg_grid)),
('EMQ', EMQ(newLR()), wrap_hyper(logreg_grid)),
('KDEy-ML', KDEyML(newLR()), {**wrap_hyper(logreg_grid), **{'bandwidth': np.logspace(-4, np.log10(0.2), 20)}}),
- ('KDEy-MLred', KDEyMLred(newLR()), {**wrap_hyper(logreg_grid), **{'bandwidth': np.logspace(-4, np.log10(0.2), 20)}}),
+ # ('KDEy-MLred', KDEyMLred(newLR()), {**wrap_hyper(logreg_grid), **{'bandwidth': np.logspace(-4, np.log10(0.2), 20)}}),
('KDEy-ML-scott', KDEyML(newLR(), bandwidth='scott'), wrap_hyper(logreg_grid)),
('KDEy-ML-silver', KDEyML(newLR(), bandwidth='silverman'), wrap_hyper(logreg_grid)),
('KDEy-ML-autoLike', KDEyMLauto2(newLR(), bandwidth='auto', target='likelihood'), wrap_hyper(logreg_grid)),
@@ -55,7 +55,7 @@ TRANSDUCTIVE_METHODS = [
#('TKDEy-ML', KDEyMLauto(newLR()), None),
# ('TKDEy-MLboth', KDEyMLauto(newLR(), optim='both'), None),
# ('TKDEy-MLbothfine', KDEyMLauto(newLR(), optim='both_fine'), None),
- # ('TKDEy-ML2', KDEyMLauto(newLR()), None),
+ ('TKDEy-ML2', KDEyMLauto(newLR(), optim='two_steps'), None),
# ('TKDEy-MLike', KDEyMLauto(newLR(), optim='max_likelihood'), None),
# ('TKDEy-MLike2', KDEyMLauto(newLR(), optim='max_likelihood2'), None),
#('TKDEy-ML3', KDEyMLauto(newLR()), None),