theorem powerMean_isFixedPoint {J : ℕ} (hJ : 0 < J) (ρ : ℝ) (hρ : ρ ≠ 0) :
IsAggFixedPoint J (powerMean J ρ hρ) := by
intro x hx
unfold aggregationOp
suffices h : (fun i => powerMean J ρ hρ (fun _j => x i)) = x by rw [h]
funext i
-- Goal: powerMean J ρ hρ (fun _j => x i) = x i
-- Case split on x i = 0 vs x i > 0
rcases eq_or_lt_of_le (hx i) with h0 | hpos
· -- x i = 0
rw [← h0]
exact powerMean_zero hJ hρ
· -- x i > 0
exact powerMean_symmetricPoint hJ hρ hposEmergence results from Paper 1, Sections 3-5: