Fix logging level setting bug

data/solverData.csv

@@ -88,3 +88,22 @@ Uni-criterion: clusterCenters,greedySwap,1000,2,10,0.13878262508660555,0.7724079
 "Uni-criterion: distinctness, distances",greedyMixed,200,2,32,0.06363425000017742,-42.57843298944509
 "Uni-criterion: distinctness, distances",greedyMixed,200,2,24,0.04308179100007692,-46.946516069915035
 "Uni-criterion: distinctness, distances",greedyMixed,200,2,93,0.19906350000019302,-75.52440490189333
+"Uni-criterion: preserveMetric, hull",greedyMinSubset,200,2,9,0.21874955005478114,0.0
+"Uni-criterion: sum, outlierness",greedySwap,200,2,40,0.07826007495168597,-90.13755098825027
+"Uni-criterion: distinctness, distances",greedySwap,200,2,60,0.37014091794844717,-79.50227671177676
+"Multi-criterion: 100*(earthMoversDistance) + 1*(distinctness, distances)",greedySwap,200,2,80,15.73197605391033,-32.55223216743509
+"Multi-criterion: 10*(earthMoversDistance) + 1*(distinctness, distances)",greedySwap,200,2,80,16.175244768033735,-76.96445453840944
+"Multi-criterion: 1*(earthMoversDistance) + 1*(distinctness, distances)",greedySwap,200,2,80,17.552692768978886,-83.85905074849741
+"Uni-criterion: distinctness, distances",greedyMixed,200,2,20,0.10706425900571048,-48.77689734061066
+"Uni-criterion: distinctness, distances",greedyMixed,200,2,34,0.17249133007135242,-58.143885051562634
+"Uni-criterion: distinctness, distances",greedyMixed,200,2,44,0.22814954188652337,-70.88814419477815
+"Uni-criterion: sum, outlierness",greedySwap,200,2,40,0.036993279005400836,-75.42053427241072
+"Uni-criterion: preserveMetric, hull",greedyMinSubset,200,2,12,0.27527309604920447,0.0
+"Uni-criterion: preserveMetric, hull",greedyMinSubset,200,2,7,0.203717328957282,0.0
+"Uni-criterion: preserveMetric, hull",greedyMinSubset,200,2,9,0.25889185711275786,0.0
+"Uni-criterion: preserveMetric, hull",greedyMinSubset,200,2,9,0.2248084848979488,0.0
+"Uni-criterion: preserveMetric, hull",greedyMinSubset,200,2,8,0.2031262150267139,0.0
+"Uni-criterion: sum, outlierness",greedySwap,200,2,40,0.035176230943761766,-66.43006587269934
+"Uni-criterion: preserveMetric, hull",greedyMinSubset,200,2,7,0.1945001189596951,0.0
+"Uni-criterion: preserveMetric, hull",greedyMinSubset,200,2,7,0.18785801599733531,0.0
+"Uni-criterion: preserveMetric, hull",greedyMinSubset,200,2,7,0.234673575963825,0.0

flexibleSubsetSelection/logger.py

@@ -23,11 +23,11 @@ def setup(name: str = "flexibleSubsetSelection",
     if not log.hasHandlers():
         handler = logging.StreamHandler(sys.stdout)
         handler.setLevel(level)
-        formatter = logging.Formatter(
-            '%(asctime)s - %(name)s - %(levelname)s: %(message)s'
-        )
+        formatString = '%(asctime)s - %(name)s - %(levelname)s: %(message)s'
+        formatter = logging.Formatter(formatString)
         handler.setFormatter(formatter)
         log.addHandler(handler)
-        log.setLevel(level)
-        log.propagate = False
+
+    log.setLevel(level)
+    log.propagate = False
     return log

flexibleSubsetSelection/objective.py

@@ -6,7 +6,6 @@
 
 # Third party
 import ot
-from ott.geometry import pointcloud
 
 import numpy as np
 from numpy.typing import ArrayLike
@@ -157,44 +156,4 @@ def entropy(array: np.ndarray) -> float:
     counts = Counter(map(tuple, array))
     total = sum(counts.values())
     probabilities = np.array(list(counts.values()))/total
-    return np.sum(probabilities * np.log(probabilities))
-
-from ott.problems.linear import linear_problem
-
-def sinkhorn(subset, fullData, solveFunction):
-    geometry = pointcloud.PointCloud(fullData, subset, epsilon=1)
-    problem = linear_problem.LinearProblem(geometry)
-    sinkhornOutput = solveFunction(problem)
-    return sinkhornOutput.reg_ot_cost
-
-
-import numpy as np
-from scipy.spatial import distance
-
-def compute_cost_matrix_chunked(dataset_array, subset_array, chunk_size=1000):
-    num_dataset = dataset_array.shape[0]
-    num_subset = subset_array.shape[0]
-    C = np.zeros((num_dataset, num_subset), dtype=np.float32)
-
-    # Compute distance in chunks
-    for i in range(0, num_dataset, chunk_size):
-        end_i = min(i + chunk_size, num_dataset)
-        C[i:end_i, :] = distance.cdist(dataset_array[i:end_i], subset_array, metric='euclidean')
-
-    return C
-
-# Compute the cost matrix in chunks
-
-def potSinkhorn(subset, fullData):
-    C = compute_cost_matrix_chunked(fullData, subset, chunk_size=1000)
-    C = C / np.max(C)
-
-    reg = 0.01  # Regularization parameter
-    a = np.ones(len(fullData)) / len(fullData)
-    b = np.ones(len(subset)) / len(subset)
-
-    transport_plan = ot.bregman.sinkhorn(a, b, C, reg, 
-                                         method="greenkhorn", 
-                                         stopThr=1e-3, 
-                                         verbose=True)
-    return np.sum(transport_plan * C)
+    return np.sum(probabilities * np.log(probabilities))