English
For ε > 0 and x ∈ ℝ, mulExpNegMulSq ε x = ε^{−1/2} · mulExpNegMulSq 1 (√ε x).
Русский
Для ε > 0 и x ∈ ℝ имеем mulExpNegMulSq ε x = ε^{−1/2} · mulExpNegMulSq 1 (√ε x).
LaTeX
$$$\\forall \\varepsilon > 0,\\; \\forall x\\in\\mathbb{R},\\; mulExpNegMulSq\\,\\varepsilon\\, x = (\\sqrt{\\varepsilon})^{-1} \\cdot mulExpNegMulSq\\,1\\, (\\sqrt{\\varepsilon}\\, x)$$$
Lean4
theorem mulExpNegMulSq_eq_sqrt_mul_mulExpNegMulSq_one (hε : 0 < ε) (x : ℝ) :
mulExpNegMulSq ε x = (√ε)⁻¹ * mulExpNegMulSq 1 (sqrt ε * x) :=
by
simp only [mulExpNegMulSq, one_mul]
have h : √ε * x * (√ε * x) = ε * x * x := by
ring_nf
rw [sq_sqrt hε.le, mul_comm]
rw [h, eq_inv_mul_iff_mul_eq₀ _]
· ring_nf
· intro h
rw [← pow_eq_zero_iff (Ne.symm (Nat.zero_ne_add_one 1)), sq_sqrt hε.le] at h
linarith