""" market_agent.py — Dynamic market environment (Factor 5: Market Demand). Simulates real-world business cycles: boom → stable → recession. Each phase affects domain demand, project availability, and profit multipliers. """ import random from src.models import MarketPhase # Per-phase domain demand multipliers (calibrated to real IT market cycles) DOMAIN_DEMAND_BY_PHASE: dict[MarketPhase, dict[str, float]] = { MarketPhase.boom: { "ai": 1.4, "web": 1.1, "mobile": 1.2, "data": 1.3, "devops": 1.0, }, MarketPhase.stable: { "ai": 1.1, "web": 1.0, "mobile": 1.0, "data": 1.0, "devops": 1.0, }, MarketPhase.recession: { "ai": 0.8, "web": 0.7, "mobile": 0.6, "data": 0.9, "devops": 1.1, # infrastructure stays essential in downturns }, } class MarketAgent: """ Markov chain market simulator. Transition probabilities modelled on typical 6-month business cycles. """ PHASES = [MarketPhase.boom, MarketPhase.stable, MarketPhase.recession] # Transition matrix: from_phase → [p(boom), p(stable), p(recession)] TRANSITION: dict[MarketPhase, list[float]] = { MarketPhase.boom: [0.30, 0.50, 0.20], MarketPhase.stable: [0.20, 0.50, 0.30], MarketPhase.recession: [0.10, 0.40, 0.50], } def __init__(self): self.phase = MarketPhase.stable def reset(self): self.phase = MarketPhase.stable def step(self) -> MarketPhase: """Advance market one quarter using Markov transition.""" weights = self.TRANSITION[self.phase] self.phase = random.choices(self.PHASES, weights=weights, k=1)[0] return self.phase def domain_demand(self) -> dict[str, float]: return DOMAIN_DEMAND_BY_PHASE[self.phase].copy() def project_count(self) -> int: """Fewer projects available in recession, more in boom.""" return { MarketPhase.boom: 4, MarketPhase.stable: 3, MarketPhase.recession: 2, }[self.phase] def profit_multiplier(self) -> float: return { MarketPhase.boom: 1.25, MarketPhase.stable: 1.00, MarketPhase.recession: 0.75, }[self.phase]