From 4e8bb05a1552b91473a9633f225672926fbc8760 Mon Sep 17 00:00:00 2001
From: Baptiste Bauvin <baptiste.bauvin@lis-lab.fr>
Date: Mon, 22 May 2023 09:58:22 -0400
Subject: [PATCH] Update for camera-ready
---
.../monoview_classifiers/gradient_boosting.py | 2 +-
.../locally_weighted_linear_regression.py | 54 +++++++++++++++++++
.../monoview_classifiers/lwlr.py | 9 ++--
.../monoview_classifiers/scmboost.py | 2 +-
.../monoview_classifiers/svm_rbf.py | 7 +--
.../monoview_classifiers/xgboost.py | 2 +-
.../multiview_platform/utils/compression.py | 17 ++++--
7 files changed, 77 insertions(+), 16 deletions(-)
diff --git a/summit/multiview_platform/monoview_classifiers/gradient_boosting.py b/summit/multiview_platform/monoview_classifiers/gradient_boosting.py
index 57288a25..281a35e9 100644
--- a/summit/multiview_platform/monoview_classifiers/gradient_boosting.py
+++ b/summit/multiview_platform/monoview_classifiers/gradient_boosting.py
@@ -44,7 +44,7 @@ class GradientBoosting(GradientBoostingClassifier, BaseMonoviewClassifier):
)
self.param_names = ["n_estimators", "max_depth", "loss", "learning_rate"]
self.classed_params = []
- self.distribs = [CustomRandint(low=50, high=500),
+ self.distribs = [CustomRandint(low=1, high=300),
CustomRandint(low=1, high=10),
['log_loss', 'deviance', 'exponential'],
CustomUniform(loc=0, state=1)]
diff --git a/summit/multiview_platform/monoview_classifiers/locally_weighted_linear_regression.py b/summit/multiview_platform/monoview_classifiers/locally_weighted_linear_regression.py
index e69de29b..23265dc4 100644
--- a/summit/multiview_platform/monoview_classifiers/locally_weighted_linear_regression.py
+++ b/summit/multiview_platform/monoview_classifiers/locally_weighted_linear_regression.py
@@ -0,0 +1,54 @@
+import numpy as np
+
+from ..monoview.monoview_utils import BaseMonoviewClassifier
+from summit.multiview_platform.utils.hyper_parameter_search import CustomUniform
+
+classifier_class_name = "LWLR"
+class LWLR(BaseMonoviewClassifier):
+
+ def __init__(self, tau=0.05, reg=0.0001, threshold=1e-6, random_state=42):
+ self.reg = reg
+ self.threshold = threshold
+ self.tau = tau
+ self.random_state = random_state
+ self.param_names = ["tau", 'reg', "threshold"]
+ self.distribs = [CustomUniform(loc=1e-2, state=1),
+ CustomUniform(loc=1e-6, state=1e-2),
+ CustomUniform(loc=1e-8, state=1e-4)]
+ self.weird_strings={}
+ self.classed_params=[]
+
+ def weights(self, x_train, x):
+ sq_diff = (x_train - x) ** 2
+ norm_sq = sq_diff.sum(axis=1)
+ return np.ravel(np.exp(- norm_sq / (2 * self.tau ** 2)))
+
+ def logistic(self, x_train):
+ return np.ravel(1 / (1 + np.exp(-x_train.dot(self.theta))))
+
+ def fit(self, X, y, **fit_params):
+ self.X = X
+ self.y = y
+
+ def train(self, x):
+ self.w = self.weights(self.X, x)
+ self.theta = np.zeros(self.X.shape[1])
+ gradient = np.ones(self.X.shape[1]) * np.inf
+ while np.linalg.norm(gradient) > self.threshold:
+ # compute gradient
+ h = self.logistic(self.X)
+ gradient = self.X.T.dot(
+ self.w * (np.ravel(self.y) - h)) - self.reg * self.theta
+ # Compute Hessian
+ D = np.diag(-(self.w * h * (1 - h)))
+ H = self.X.T.dot(D).dot(self.X) - self.reg * np.identity(
+ self.X.shape[1])
+ # weight update
+ self.theta = self.theta - np.linalg.inv(H).dot(gradient)
+
+ def predict(self, X):
+ preds = []
+ for x in X:
+ self.train(x)
+ preds.append(np.array(self.logistic(X) > 0.5).astype(int)[0])
+ return np.array(preds)
diff --git a/summit/multiview_platform/monoview_classifiers/lwlr.py b/summit/multiview_platform/monoview_classifiers/lwlr.py
index 48db51d3..2387ed17 100644
--- a/summit/multiview_platform/monoview_classifiers/lwlr.py
+++ b/summit/multiview_platform/monoview_classifiers/lwlr.py
@@ -1,5 +1,4 @@
-from summit.multiview_platform.monoview_classifiers.additions.SVCClassifier import \
- SVCClassifier
+from learners.algorithms.lwlr import LWLRLearner
from ..monoview.monoview_utils import BaseMonoviewClassifier
from summit.multiview_platform.utils.hyper_parameter_search import CustomUniform
@@ -11,15 +10,15 @@ __status__ = "Prototype" # Production, Development, Prototype
classifier_class_name = "SVMRBF"
-class SVMRBF(SVCClassifier, BaseMonoviewClassifier):
+class LWLRClassifier(LWLRLearner, BaseMonoviewClassifier):
"""
This class is an adaptation of scikit-learn's `SVC <https://scikit-learn.org/stable/modules/generated/sklearn.svm.SVC.html>`_
Here, it is the RBF kernel version
"""
- def __init__(self, random_state=None, gamma="auto", C=1.0, **kwargs):
- SVCClassifier.__init__(self,
+ def __init__(self, random_state=None, sigma="auto", nn=1.0, **kwargs):
+ LWLRLearner.__init__(self,
C=C,
kernel='rbf',
gamma=gamma,
diff --git a/summit/multiview_platform/monoview_classifiers/scmboost.py b/summit/multiview_platform/monoview_classifiers/scmboost.py
index 89f1e4b7..070c1207 100644
--- a/summit/multiview_platform/monoview_classifiers/scmboost.py
+++ b/summit/multiview_platform/monoview_classifiers/scmboost.py
@@ -40,7 +40,7 @@ class SCMboost(AdaBoostClassifier, BaseMonoviewClassifier):
algorithm="SAMME",)
self.param_names = ["n_estimators", "base_estimator__p", "base_estimator__model_type", "base_estimator__max_rules"]
self.classed_params = []
- self.distribs = [CustomRandint(low=1, high=100), CustomUniform(loc=0, state=1), ["conjunction", "disjunction"], CustomRandint(low=1, high=20)]
+ self.distribs = [CustomRandint(low=1, high=100), CustomUniform(loc=0, state=1), ["conjunction", "disjunction"], CustomRandint(low=1, high=5)]
self.weird_strings = {}
diff --git a/summit/multiview_platform/monoview_classifiers/svm_rbf.py b/summit/multiview_platform/monoview_classifiers/svm_rbf.py
index 504ef095..48db51d3 100644
--- a/summit/multiview_platform/monoview_classifiers/svm_rbf.py
+++ b/summit/multiview_platform/monoview_classifiers/svm_rbf.py
@@ -18,11 +18,12 @@ class SVMRBF(SVCClassifier, BaseMonoviewClassifier):
Here, it is the RBF kernel version
"""
- def __init__(self, random_state=None, C=1.0, **kwargs):
+ def __init__(self, random_state=None, gamma="auto", C=1.0, **kwargs):
SVCClassifier.__init__(self,
C=C,
kernel='rbf',
+ gamma=gamma,
random_state=random_state
)
- self.param_names = ["C", "random_state"]
- self.distribs = [CustomUniform(loc=0, state=1), [random_state]]
+ self.param_names = ["C", 'gamma', "random_state"]
+ self.distribs = [CustomUniform(loc=1e-3, state=1e3), CustomUniform(loc=1e-1, state=1e1), [random_state]]
diff --git a/summit/multiview_platform/monoview_classifiers/xgboost.py b/summit/multiview_platform/monoview_classifiers/xgboost.py
index 2b44f211..280b1150 100644
--- a/summit/multiview_platform/monoview_classifiers/xgboost.py
+++ b/summit/multiview_platform/monoview_classifiers/xgboost.py
@@ -32,7 +32,7 @@ class XGB(XGBClassifier, BaseMonoviewClassifier):
random_state=random_state)
self.param_names = ["n_estimators", "learning_rate", "max_depth", "objective"]
self.classed_params = []
- self.distribs = [CustomRandint(low=10, high=500),
+ self.distribs = [CustomRandint(low=1, high=300),
CustomUniform(),
CustomRandint(low=1, high=10),
['binary:logistic', 'binary:hinge', ],]
diff --git a/summit/multiview_platform/utils/compression.py b/summit/multiview_platform/utils/compression.py
index 6fb7cb49..a2298f27 100644
--- a/summit/multiview_platform/utils/compression.py
+++ b/summit/multiview_platform/utils/compression.py
@@ -43,14 +43,21 @@ def remove_compressed(exp_path):
if __name__=="__main__":
- # for dir in os.listdir("/home/baptiste/Documents/Gitwork/summit/results/"):
+ for dir in os.listdir("/home/baptiste/Documents/Gitwork/summit/results/"):
+ if os.path.isdir(os.path.join("/home/baptiste/Documents/Gitwork/summit/results/", dir)):
+ for exp in os.listdir((os.path.join("/home/baptiste/Documents/Gitwork/summit/results/", dir))):
+ print("\t", exp)
+ if os.path.isdir(os.path.join("/home/baptiste/Documents/Gitwork/summit/results/", dir, exp)):
+ explore_files(os.path.join("/home/baptiste/Documents/Gitwork/summit/results/", dir, exp))
+ plif = dict()
+ # for dir in os.listdir("/home/baptiste/Documents/Clouded/short_projects/SCMBoost/results"):
# print(dir)
- # for exp in os.listdir((os.path.join("/home/baptiste/Documents/Gitwork/summit/results/", dir))):
+ # for exp in os.listdir((os.path.join("/home/baptiste/Documents/Clouded/short_projects/SCMBoost/results", dir))):
# print("\t", exp)
- # if os.path.isdir(os.path.join("/home/baptiste/Documents/Gitwork/summit/results/", dir, exp)):
- # explore_files(os.path.join("/home/baptiste/Documents/Gitwork/summit/results/", dir, exp))
+ # if os.path.isdir(os.path.join("/home/baptiste/Documents/Clouded/short_projects/SCMBoost/results", dir, exp)):
+ # explore_files(os.path.join("/home/baptiste/Documents/Clouded/short_projects/SCMBoost/results", dir, exp))
# # explore_files("/home/baptiste/Documents/Gitwork/biobanq_covid_expes/results/")
- explore_files("/home/baptiste/Documents/Gitwork/summit/results/clinical/debug_started_2023_04_05-08_23_00_bal_acc")
+ # explore_files("/home/baptiste/Documents/Gitwork/summit/results/clinical/debug_started_2023_04_05-08_23_00_bal_acc")
# explore_files(
# "/home/baptiste/Documents/Gitwork/summit/results/lives_thesis_EMF/debug_started_2023_03_24-10_02_21_thesis")
# # simplify_plotly("/home//baptiste/Documents/Gitwork/summit/results/hepatitis/debug_started_2022_03_16-15_06_55__/hepatitis-mean_on_10_iter-balanced_accuracy_p.html")
--
GitLab