OAI/LinearAlgebra/MatrixMultiplication/AuxiliarySeparation/Growth/Floor.lean

GitHub source file

4 declarations, 0 contain sorry, 0 declared as axioms

Growth rates of floor-rounded exponents

The integer exponents obtained by rounding a nonnegative multiple of the tensor power have the intended normalized growth rate. Applying a positive base to these exponents preserves that rate after taking roots.

  1. theorem OAI.MatrixMultiplication.AuxiliarySeparation.tendsto_floor_mul_div line 21

    Source documentation

    Rounding a nonnegative linear exponent down has vanishing normalized error.

    theorem tendsto_floor_mul_div {c : ℝ} (hc : 0 ≤ c) :
        Tendsto (fun j : ℕ => (⌊(j : ℝ) * c⌋₊ : ℝ) / (j : ℝ))
          atTop (𝓝 c) := by
    …
  2. theorem OAI.MatrixMultiplication.AuxiliarySeparation.tendsto_floor_mul_atTop line 29

    Source documentation

    A floor-rounded exponent diverges when its linear coefficient is positive.

    theorem tendsto_floor_mul_atTop {c : ℝ} (hc : 0 < c) :
        Tendsto (fun j : ℕ => ⌊(j : ℝ) * c⌋₊) atTop atTop := by
    …
  3. theorem OAI.MatrixMultiplication.AuxiliarySeparation.tendsto_pow_floor_rpow_inv line 34

    Source documentation

    The root growth rate of powers with floor-rounded exponents.

    theorem tendsto_pow_floor_rpow_inv {u c : ℝ} (hu : 0 < u) (hc : 0 ≤ c) :
        Tendsto (fun j : ℕ => (u ^ ⌊(j : ℝ) * c⌋₊) ^ (1 / (j : ℝ)))
          atTop (𝓝 (u ^ c)) := by
    …
  4. theorem OAI.MatrixMultiplication.AuxiliarySeparation.tendsto_pow_floor_logb_div_rpow_inv line 46

    Source documentation

    The logarithmic choice of exponent gives the prescribed root growth rate. In the tensor application, s is the exponent bound with a positive slack.

    theorem tendsto_pow_floor_logb_div_rpow_inv {u n s : ℝ}
        (hu : 1 < u) (hn : 1 ≤ n) (hs : 0 < s) :
        Tendsto
          (fun j : ℕ => (u ^ ⌊(j : ℝ) * (Real.logb u n / s)⌋₊) ^ (1 / (j : ℝ)))
          atTop (𝓝 (n ^ (1 / s))) := by
    …