approval validation

szufix · szufix · commit 0af82b276f18 · 2024-01-30T17:59:49.000+01:00
diff --git a/prefsampling/filters/approval_filters.py b/prefsampling/filters/approval_filters.py
@@ -25,7 +25,9 @@ def permute_approval_voters(votes: list[set[int]], seed: int = None) -> list[set
 
 
 def rename_approval_candidates(
-    votes: list[set[int]], seed: int = None
+        votes: list[set[int]],
+        seed: int = None,
+        num_candidates: int = None
 ) -> list[set[int]]:
     """
     Renames the candidates in approval votes.
@@ -36,18 +38,20 @@ def rename_approval_candidates(
             Approval votes.
         seed : int
             Seed for numpy random number generator.
+        num_candidates : int
+            Number of Candidates.
 
     Returns
     -------
         list[set[int]]
             Approval votes.
     """
-    rng = np.random.default_rng(seed)
-    max_id = max([max(vote) for vote in votes if len(vote) > 0])
-    mapping = rng.permutation(max_id + 1)
 
+    rng = np.random.default_rng(seed)
+    if num_candidates is None:
+        num_candidates = max([max(vote) for vote in votes if len(vote) > 0]) + 1
+    mapping = rng.permutation(num_candidates)
     votes = [{mapping[c] for c in vote} for vote in votes]
-
     return votes
 
 
diff --git a/prefsampling/ordinal/singlecrossing.py b/prefsampling/ordinal/singlecrossing.py
@@ -85,9 +85,9 @@ def single_crossing(
         votes[i, :] = domain[index]
         last_sampled_index = index
 
-    # vote_indices = np.sort(rng.choice(np.arange(domain_size), size=num_voters))
-    # for i, index in enumerate(vote_indices):
-    #     votes[i, :] = domain[index]
+    vote_indices = np.sort(rng.choice(np.arange(domain_size), size=num_voters))
+    for i, index in enumerate(vote_indices):
+        votes[i, :] = domain[index]
     return votes
 
 
diff --git a/prefsampling/tree/__init__.py b/prefsampling/tree/__init__.py
@@ -13,4 +13,6 @@
     "schroeder_tree_lescanne",
     "schroeder_tree_brute_force",
     "all_schroeder_tree",
+    "caterpillar_tree",
+    "caterpillar_tree",
 ]
diff --git a/validation/approval/__init__.py b/validation/approval/__init__.py
diff --git a/validation/approval/euclidean.py b/validation/approval/euclidean.py
@@ -0,0 +1,64 @@
+from prefsampling.core.euclidean import EuclideanSpace
+from prefsampling.ordinal import euclidean
+from validation.utils import get_all_subsets
+from validation.validator import Validator
+
+
+class ApprovalEuclideanValidatorUniform(Validator):
+    def __init__(self):
+        parameters_list = []
+        for space in EuclideanSpace:
+            for dimension in [2, 3]:
+                parameters_list.append(
+                    {
+                        "num_voters": 50,
+                        "num_candidates": 5,
+                        "space": space,
+                        "dimension": dimension,
+                    },
+                )
+        super(ApprovalEuclideanValidatorUniform, self).__init__(
+            parameters_list,
+            "Euclidean",
+            "euclidean_uniform",
+            True,
+            sampler_func=euclidean,
+            constant_parameters=("num_voters", "num_candidates"),
+            faceted_parameters=("space", "dimension"),
+        )
+
+    def sample_cast(self, sample):
+        return tuple(sample[0])
+
+    def all_outcomes(self, sampler_parameters):
+        return get_all_ranks(sampler_parameters["num_candidates"])
+
+    def theoretical_distribution(self, sampler_parameters, all_outcomes) -> dict:
+        return {o: 1 / len(all_outcomes) for o in all_outcomes}
+
+
+class OrdinalEuclideanValidator(Validator):
+    def __init__(self):
+        parameters_list = []
+        for space in EuclideanSpace:
+            for dimension in [2, 3]:
+                parameters_list.append(
+                    {
+                        "num_voters": 3,
+                        "num_candidates": 3,
+                        "space": space,
+                        "dimension": dimension,
+                    },
+                )
+        super(OrdinalEuclideanValidator, self).__init__(
+            parameters_list,
+            "Euclidean",
+            "euclidean",
+            False,
+            sampler_func=euclidean,
+            constant_parameters=("num_voters", "num_candidates"),
+            faceted_parameters=("space", "dimension"),
+        )
+
+    def sample_cast(self, sample):
+        return tuple(tuple(r) for r in sample)
diff --git a/validation/approval/identity.py b/validation/approval/identity.py
@@ -0,0 +1,30 @@
+from prefsampling.approval import identity
+from validation.utils import get_all_subsets
+from validation.validator import Validator
+
+
+class ApprovalIdentityValidator(Validator):
+    def __init__(self):
+        parameters_list = [
+            {"num_voters": 1, "num_candidates": 4, "p": 0.25},
+            {"num_voters": 1, "num_candidates": 4, "p": 0.5},
+        ]
+        super(ApprovalIdentityValidator, self).__init__(
+            parameters_list,
+            "Identity",
+            "identity",
+            True,
+            sampler_func=identity,
+            constant_parameters=("num_voters", "num_candidates"),
+            faceted_parameters="p",
+        )
+
+    def all_outcomes(self, sampler_parameters):
+        return get_all_subsets(sampler_parameters["num_candidates"])
+
+    def theoretical_distribution(self, sampler_parameters, all_outcomes) -> dict:
+        k = int(sampler_parameters["p"] * sampler_parameters["num_candidates"])
+        return {str(k): 1}
+
+    def sample_cast(self, sample):
+        return str(len(sample[0]))
diff --git a/validation/approval/impartial.py b/validation/approval/impartial.py
@@ -0,0 +1,44 @@
+from prefsampling.approval import impartial
+from validation.utils import get_all_subsets
+from validation.validator import Validator
+from math import factorial
+
+
+class ApprovalImpartialValidator(Validator):
+    def __init__(self):
+        parameters_list = [
+            {"num_voters": 1, "num_candidates": 4, "p": 0.3},
+            {"num_voters": 1, "num_candidates": 4, "p": 0.5},
+            {"num_voters": 1, "num_candidates": 4, "p": 0.7},
+            {"num_voters": 1, "num_candidates": 5, "p": 0.3},
+            {"num_voters": 1, "num_candidates": 5, "p": 0.5},
+            {"num_voters": 1, "num_candidates": 5, "p": 0.7},
+        ]
+        super(ApprovalImpartialValidator, self).__init__(
+            parameters_list,
+            "Impartial",
+            "impartial_apr",
+            True,
+            sampler_func=impartial,
+            constant_parameters="num_voters",
+            faceted_parameters=("p", "num_candidates"),
+        )
+
+    def all_outcomes(self, sampler_parameters):
+        return get_all_subsets(sampler_parameters["num_candidates"])
+
+    def theoretical_distribution(self, sampler_parameters, all_outcomes) -> dict:
+
+        m = sampler_parameters["num_candidates"]
+        p = sampler_parameters["p"]
+
+        probabilities = [None for _ in range(m+1)]
+        for k in range(m+1):
+            binomial_coefficient = factorial(m) / (factorial(k) * factorial(m - k))
+            probability = binomial_coefficient * (p ** k) * ((1 - p) ** (m - k))
+            probabilities[k] = probability
+
+        return {str(len(o)): probabilities[len(o)] for o in all_outcomes}
+
+    def sample_cast(self, sample):
+        return str(len(sample[0]))
diff --git a/validation/approval/noise.py b/validation/approval/noise.py
@@ -0,0 +1,42 @@
+import math
+
+from prefsampling.approval import noise
+from validation.utils import get_all_subsets, hamming
+from validation.validator import Validator
+
+
+class ApprovalNoiseValidator(Validator):
+    def __init__(self):
+        parameters_list = [
+            {"num_voters": 1, "num_candidates": 4, "phi": 0.25, "p": 0.25},
+            {"num_voters": 1, "num_candidates": 4, "phi": 0.25, "p": 0.5},
+            {"num_voters": 1, "num_candidates": 4, "phi": 0.25, "p": 0.75},
+            {"num_voters": 1, "num_candidates": 4, "phi": 0.5, "p": 0.25},
+            {"num_voters": 1, "num_candidates": 4, "phi": 0.5, "p": 0.5},
+            {"num_voters": 1, "num_candidates": 4, "phi": 0.5, "p": 0.75},
+        ]
+        super(ApprovalNoiseValidator, self).__init__(
+            parameters_list,
+            "Noise",
+            "noise",
+            True,
+            sampler_func=noise,
+            constant_parameters=("num_voters", "num_candidates"),
+            faceted_parameters=("phi", "p"),
+        )
+
+    def all_outcomes(self, sampler_parameters):
+        return get_all_subsets(sampler_parameters["num_candidates"])
+
+    def theoretical_distribution(self, sampler_parameters, all_outcomes) -> dict:
+        m = sampler_parameters["num_candidates"]
+        p = sampler_parameters["p"]
+        phi = sampler_parameters["phi"]
+        k = math.floor(p*m)
+        central_vote = {i for i in range(k)}
+        tmp_dict = {str(o): phi**hamming(central_vote, o) for o in all_outcomes}
+        denom = sum(tmp_dict.values())
+        return {str(o): tmp_dict[str(o)]/denom for o in all_outcomes}
+
+    def sample_cast(self, sample):
+        return str(sample[0])
diff --git a/validation/approval/partylist.py b/validation/approval/partylist.py
@@ -0,0 +1,30 @@
+from prefsampling.approval import partylist
+# from validation.utils import get_all_ranks, get_all_anonymous_profiles
+from validation.validator import Validator
+
+
+class ApprovalImpartialValidator(Validator):
+    def __init__(self):
+        parameters_list = [
+            {"num_voters": 1, "num_candidates": 4},
+            {"num_voters": 1, "num_candidates": 5},
+            {"num_voters": 1, "num_candidates": 6},
+        ]
+        super(ApprovalImpartialValidator, self).__init__(
+            parameters_list,
+            "Partylist",
+            "partylist",
+            True,
+            sampler_func=partylist,
+            constant_parameters="num_voters",
+            faceted_parameters="num_candidates",
+        )
+
+    # def all_outcomes(self, sampler_parameters):
+    #     return get_all_ranks(sampler_parameters["num_candidates"])
+
+    # def theoretical_distribution(self, sampler_parameters, all_outcomes) -> dict:
+    #     return {o: 1 / len(all_outcomes) for o in all_outcomes}
+
+    # def sample_cast(self, sample):
+    #     return tuple(sample[0])
diff --git a/validation/approval/resampling.py b/validation/approval/resampling.py
@@ -0,0 +1,55 @@
+import math
+
+from prefsampling.approval import resampling
+from validation.utils import get_all_subsets
+from validation.validator import Validator
+
+
+class ApprovalResamplingValidator(Validator):
+    def __init__(self):
+        parameters_list = [
+            {"num_voters": 1, "num_candidates": 6, "phi": 0.25, "p": 0.5},
+            {"num_voters": 1, "num_candidates": 6, "phi": 0.5, "p": 0.5},
+            {"num_voters": 1, "num_candidates": 6, "phi": 0.75, "p": 0.5},
+            {"num_voters": 1, "num_candidates": 6, "phi": 0.25, "p": 0.34},
+            {"num_voters": 1, "num_candidates": 6, "phi": 0.5, "p": 0.34},
+            {"num_voters": 1, "num_candidates": 6, "phi": 0.75, "p": 0.34},
+        ]
+        super(ApprovalResamplingValidator, self).__init__(
+            parameters_list,
+            "Resampling",
+            "resampling",
+            True,
+            sampler_func=resampling,
+            constant_parameters=("num_voters", "num_candidates"),
+            faceted_parameters=("phi", "p"),
+        )
+
+    def all_outcomes(self, sampler_parameters):
+        return get_all_subsets(sampler_parameters["num_candidates"])
+
+    def theoretical_distribution(self, sampler_parameters, all_outcomes) -> dict:
+        m = sampler_parameters["num_candidates"]
+        p = sampler_parameters["p"]
+        phi = sampler_parameters["phi"]
+        k = math.floor(p*m)
+        central_vote = {i for i in range(k)}
+
+        A = {}
+        for outcome in all_outcomes:
+            prob = 1
+            for c in range(m):
+                if c in central_vote and c in outcome:
+                    prob *= (1-phi) + phi*p
+                elif c in central_vote and c not in outcome:
+                    prob *= phi*(1-p)
+                elif c not in central_vote and c in outcome:
+                    prob *= phi*p
+                else:
+                    prob *= (1-phi) + phi*(1-p)
+            A[str(outcome)] = prob
+        return A
+
+    def sample_cast(self, sample):
+        return str(sample[0])
+
diff --git a/validation/approval/truncated_urn.py b/validation/approval/truncated_urn.py
@@ -0,0 +1,30 @@
+from prefsampling.approval import impartial
+from validation.utils import get_all_subsets
+from validation.validator import Validator
+
+
+class ApprovalImpartialValidator(Validator):
+    def __init__(self):
+        parameters_list = [
+            {"num_voters": 1, "num_candidates": 4},
+            {"num_voters": 1, "num_candidates": 5},
+            {"num_voters": 1, "num_candidates": 6},
+        ]
+        super(ApprovalImpartialValidator, self).__init__(
+            parameters_list,
+            "Impartial",
+            "impartial",
+            True,
+            sampler_func=impartial,
+            constant_parameters="num_voters",
+            faceted_parameters="num_candidates",
+        )
+
+    def all_outcomes(self, sampler_parameters):
+        return get_all_subsets(sampler_parameters["num_candidates"])
+
+    # def theoretical_distribution(self, sampler_parameters, all_outcomes) -> dict:
+    #     return {o: 1 / len(all_outcomes) for o in all_outcomes}
+
+    # def sample_cast(self, sample):
+    #     return tuple(sample[0])
diff --git a/validation/csv/ordinal/impartial.csv b/validation/csv/ordinal/impartial.csv
diff --git a/validation/run.py b/validation/run.py
diff --git a/validation/utils.py b/validation/utils.py
diff --git a/validation/validator.py b/validation/validator.py

Original file line number	Diff line number	Diff line change
`@@ -13,4 +13,6 @@`
`13`	`13`	`"schroeder_tree_lescanne",`
`14`	`14`	`"schroeder_tree_brute_force",`
`15`	`15`	`"all_schroeder_tree",`
	`16`	`+ "caterpillar_tree",`
	`17`	`+ "caterpillar_tree",`
`16`	`18`	`]`