File size: 7,875 Bytes
484e3bc |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 |
"""
Scenario representation and management for geopolitical modeling.
"""
import numpy as np
from typing import Dict, List, Optional, Any
from dataclasses import dataclass, field
from datetime import datetime
@dataclass
class Scenario:
"""
Represents a geopolitical scenario with multiple features and metadata.
Attributes
----------
name : str
Name or identifier for the scenario
features : Dict[str, np.ndarray]
Dictionary of feature names to values
timestamp : datetime
Timestamp of the scenario
metadata : Dict[str, Any]
Additional metadata
probability : float
Probability or weight of this scenario (for ensembles)
"""
name: str
features: Dict[str, np.ndarray]
timestamp: datetime = field(default_factory=datetime.now)
metadata: Dict[str, Any] = field(default_factory=dict)
probability: float = 1.0
def get_feature_vector(self, feature_names: Optional[List[str]] = None) -> np.ndarray:
"""
Get features as a vector.
Parameters
----------
feature_names : List[str], optional
List of feature names to include (if None, use all)
Returns
-------
np.ndarray
Feature vector
"""
if feature_names is None:
feature_names = list(self.features.keys())
vectors = [self.features[name].flatten() for name in feature_names if name in self.features]
return np.concatenate(vectors)
def get_feature_matrix(self) -> np.ndarray:
"""
Get all features as a matrix.
Returns
-------
np.ndarray
Feature matrix (n_features, ...)
"""
return np.array([v for v in self.features.values()])
def add_feature(self, name: str, values: np.ndarray) -> None:
"""
Add a new feature to the scenario.
Parameters
----------
name : str
Feature name
values : np.ndarray
Feature values
"""
self.features[name] = values
def remove_feature(self, name: str) -> None:
"""
Remove a feature from the scenario.
Parameters
----------
name : str
Feature name to remove
"""
if name in self.features:
del self.features[name]
def clone(self) -> 'Scenario':
"""
Create a deep copy of the scenario.
Returns
-------
Scenario
Cloned scenario
"""
return Scenario(
name=self.name,
features={k: v.copy() for k, v in self.features.items()},
timestamp=self.timestamp,
metadata=self.metadata.copy(),
probability=self.probability
)
class ScenarioDistribution:
"""
Represents a distribution over multiple scenarios.
This is useful for Monte Carlo simulations, ensemble forecasting,
and probabilistic reasoning.
"""
def __init__(self, scenarios: Optional[List[Scenario]] = None):
"""
Initialize scenario distribution.
Parameters
----------
scenarios : List[Scenario], optional
Initial list of scenarios
"""
self.scenarios: List[Scenario] = scenarios if scenarios is not None else []
def add_scenario(self, scenario: Scenario) -> None:
"""
Add a scenario to the distribution.
Parameters
----------
scenario : Scenario
Scenario to add
"""
self.scenarios.append(scenario)
def get_probabilities(self) -> np.ndarray:
"""
Get probabilities of all scenarios.
Returns
-------
np.ndarray
Array of probabilities
"""
probs = np.array([s.probability for s in self.scenarios])
# Normalize
return probs / probs.sum()
def normalize_probabilities(self) -> None:
"""
Normalize scenario probabilities to sum to 1.
"""
total_prob = sum(s.probability for s in self.scenarios)
for scenario in self.scenarios:
scenario.probability /= total_prob
def get_feature_samples(self, feature_names: Optional[List[str]] = None) -> np.ndarray:
"""
Get feature samples from all scenarios.
Parameters
----------
feature_names : List[str], optional
List of feature names to include
Returns
-------
np.ndarray
Feature samples (n_scenarios, n_features)
"""
samples = [s.get_feature_vector(feature_names) for s in self.scenarios]
return np.array(samples)
def get_weighted_mean(self, feature_names: Optional[List[str]] = None) -> np.ndarray:
"""
Compute weighted mean of features.
Parameters
----------
feature_names : List[str], optional
List of feature names to include
Returns
-------
np.ndarray
Weighted mean feature vector
"""
samples = self.get_feature_samples(feature_names)
probs = self.get_probabilities()
return np.average(samples, axis=0, weights=probs)
def get_variance(self, feature_names: Optional[List[str]] = None) -> np.ndarray:
"""
Compute variance of features.
Parameters
----------
feature_names : List[str], optional
List of feature names to include
Returns
-------
np.ndarray
Variance of features
"""
samples = self.get_feature_samples(feature_names)
probs = self.get_probabilities()
mean = self.get_weighted_mean(feature_names)
variance = np.average((samples - mean) ** 2, axis=0, weights=probs)
return variance
def sample(self, n_samples: int = 1, replace: bool = True) -> List[Scenario]:
"""
Sample scenarios from the distribution.
Parameters
----------
n_samples : int
Number of samples to draw
replace : bool
Whether to sample with replacement
Returns
-------
List[Scenario]
Sampled scenarios
"""
probs = self.get_probabilities()
indices = np.random.choice(
len(self.scenarios),
size=n_samples,
replace=replace,
p=probs
)
return [self.scenarios[i] for i in indices]
def filter_by_probability(self, threshold: float) -> 'ScenarioDistribution':
"""
Filter scenarios by probability threshold.
Parameters
----------
threshold : float
Minimum probability threshold
Returns
-------
ScenarioDistribution
New distribution with filtered scenarios
"""
filtered_scenarios = [s for s in self.scenarios if s.probability >= threshold]
return ScenarioDistribution(filtered_scenarios)
def get_top_k(self, k: int) -> 'ScenarioDistribution':
"""
Get top k scenarios by probability.
Parameters
----------
k : int
Number of scenarios to return
Returns
-------
ScenarioDistribution
Distribution with top k scenarios
"""
sorted_scenarios = sorted(self.scenarios, key=lambda s: s.probability, reverse=True)
return ScenarioDistribution(sorted_scenarios[:k])
def __len__(self) -> int:
"""Return number of scenarios."""
return len(self.scenarios)
def __getitem__(self, idx: int) -> Scenario:
"""Get scenario by index."""
return self.scenarios[idx]
def __iter__(self):
"""Iterate over scenarios."""
return iter(self.scenarios)
|