wild_triple_partition_mass_le_of_odd_ge_179
For every odd number q of triple blocks with q>=179, the existing sufficient canonical seal-event resource sum wtc_mass(q) is at most1/50. This is an unproved arithmetic helper candidate; it asserts no main-conjecture conclusion.
Statement
theorem wild_triple_partition_mass_le_of_odd_ge_179
(q : ℕ) (hq : 179 ≤ q) (hodd : q%2 = 1) :
wtc_mass q ≤ (1/50 : ℝ)- Source
- IN-359 candidate continuation of the existing OP-306 permutation resource criterion. Exact audits8965/8967 locate a finite baseline; uniform monotonicity and tail summation must still be proved. It is not an accepted improved wild-number theorem or a research advance while stated.
- Verified
- 02 Nov 2026
- Axioms
Classical.choiceQuot.soundpropext- Built on 5
- wild_triple_partition_large_order_mass_boundFor odd q>=100001, both canonical seal-event resource sums for q triples of vertices in K_(3q), with weight (21/20)^d/(3q)_d, are at most…
- wild_triple_partition_term_geometric_tailFor q>=3 odd and any h>=5 with2*h<=q, the existing canonical triple-partition resource term wtc_term(q,h) is at most 3*h*(27/40)^h. The…
- wild_triple_partition_term_antitone_odd_stepFor every odd q>=3 and admissible fixed cut size h>=1 with2*h<=q, the canonical triple-partition resource term wtc_term(q,h) weakly…
- wild_triple_partition_mass_179_baselineThe sum of the canonical triple-partition resource terms wtc_term(179, h) for cut sizes h from 1 to 39 is at most 989/50000 = 0.01978.
- g$g(S,M) = \dim_{\mathbb{F}_p}(S^{\perp}_M) - \dim_{\mathbb{F}_p}(S_M)$.
The proof, explained
The proof decomposes the total mass wtc_mass(q) = ∑_{h=1}^{⌊q/2⌋} wtc_term(q, h) for an odd integer q ≥ 179 into a head sum (h from 1 to 39) and a tail sum (h from 40 to ⌊q/2⌋). The head sum is bounded by 989/50000 by establishing that wtc_term(179+2k, h) ≤ wtc_term(179, h) and using a precomputed baseline bound at q = 179. The tail sum is bounded by 1/10000 by dominating the terms with an arithmetic-geometric sequence 3*h*(27/40)^h and bounding the series using a closed-form telescoping identity. Summing the two bounds gives 989/50000 + 1/10000 = 497/25000 ≤ 1/50.
For any sequence f : ℕ → ℝ and a ≤ b, the telescoping sum ∑_{k=a}^{b-1} (f(k) - f(k+1)) = f(a) - f(b).
We proceed by induction on b ≥ a. In the base case b = a, the sum is over an empty interval and equals 0 = f(a) - f(a). In the induction step, we peel off the top term f(b') - f(b'+1) from the sum over [a, b'+1) using Finset.sum_Ico_succ_top and apply the induction hypothesis; ring arithmetic simplifies (f(a) - f(b')) + (f(b') - f(b'+1)) to f(a) - f(b'+1).
uses
Finset.sum_Ico_succ_topinduction' h with b' hb' ih · simp · rw [Finset.sum_Ico_succ_top hb', ih] ring
For g(n) = (n/(1-r) + r/(1-r)^2) * r^n with 0 ≤ r < 1, each term satisfies k * r^k = g(k) - g(k+1).
Expanding the definition of g, factoring r^(k+1) as r^k * r via pow_succ, and casting k+1 to the reals, the difference g(k) - g(k+1) factors as ((k/(1-r) + r/(1-r)^2) - ((k+1)/(1-r) + r/(1-r)^2)*r) * r^k. Clearing denominators with field_simp and expanding via ring confirms that the bracketed difference simplifies to k.
uses
pow_succintro g have hdiff : ∀ k : ℕ, (k : ℝ) * r^k = g k - g (k + 1) := by intro k dsimp [g] rw [pow_succ r k] have h1 : ((k + 1 : ℕ) : ℝ) = (k : ℝ) + 1 := by push_cast; rfl rw [h1] have hr : 1 - r ≠ 0 := by linarith have h3 : ((k : ℝ) / (1 - r) + r / (1 - r)^2) - (((k : ℝ) + 1) / (1 - r) + r / (1 - r)^2) * r = (k : ℝ) := by field_simp ring calc (k : ℝ) * r^k = (((k : ℝ) / (1 - r) + r / (1 - r)^2) - (((k : ℝ) + 1) / (1 - r) + r / (1 - r)^2) * r) * r^k := by rw [h3] _ = ((k : ℝ) / (1 - r) + r / (1 - r)^2) * r^k - (((k : ℝ) + 1) / (1 - r) + r / (1 - r)^2) * (r^k * r) := by ring
The sum ∑_{k=a}^{b-1} k * r^k equals g(a) - g(b).
We replace each summand k * r^k with g(k) - g(k+1) using sum congruence, then apply the telescoping sum lemma to evaluate the resulting sum as g(a) - g(b).
uses
Finset.sum_congr,wild_triple_partition_mass_le_of_odd_ge_179_sum_Ico_telescope_subhave hcongr : ∑ k ∈ Finset.Ico a b, (k : ℝ) * r^k = ∑ k ∈ Finset.Ico a b, (g k - g (k + 1)) := by apply Finset.sum_congr rfl; intro k _; exact hdiff k rw [hcongr, wild_triple_partition_mass_le_of_odd_ge_179_sum_Ico_telescope_sub g a b hab]
The sum ∑_{k=a}^{b-1} k * r^k is at most g(a).
Because b ≥ 0, r ≥ 0, and 1 - r > 0, the terms b/(1-r) and r/(1-r)^2 are non-negative, and r^b ≥ 0. Hence g(b) ≥ 0, which immediately implies g(a) - g(b) ≤ g(a).
uses
div_nonneg,Nat.cast_nonneg,pow_nonneg,mul_nonneg,add_nonneghave hgb : 0 ≤ g b := by dsimp [g] have hb1 : 0 ≤ (b:ℝ)/(1-r) := div_nonneg (Nat.cast_nonneg b) (by linarith) have hb2 : 0 ≤ r/(1-r)^2 := div_nonneg hr0 (pow_nonneg (by linarith) 2) exact mul_nonneg (add_nonneg hb1 hb2) (pow_nonneg hr0 b) linarith
For 1 ≤ h with 2*h ≤ 179 and any k ∈ ℕ, wtc_term(179 + 2*k, h) ≤ wtc_term(179, h).
We induct on k. The base case k = 0 is trivial. For the step k+1, because 179 + 2*k is an odd integer at least 3, the step monotonicity lemma wild_triple_partition_term_antitone_odd_step gives wtc_term(179 + 2*k + 2, h) ≤ wtc_term(179 + 2*k, h). Transitivity with the induction hypothesis completes the step.
uses
wild_triple_partition_term_antitone_odd_step,le_transintro k induction k with | zero => simp | succ k ih => have hmono : wtc_term (179 + 2*k + 2) h ≤ wtc_term (179 + 2*k) h := wild_triple_partition_term_antitone_odd_step (179 + 2*k) h (by omega) (by omega) hh (by omega) have heq : 179 + 2*(k+1) = 179 + 2*k + 2 := by ring rw [heq] exact le_trans hmono ih
The total mass wtc_mass(q) equals the sum ∑_{h=1}^{⌊q/2⌋} wtc_term(q, h), expressed as the half-open interval sum over h ∈ [1, q/2 + 1).
By definition, wtc_mass(q) is the sum over range(q+1) filtered by 1 ≤ h and h ≤ q/2. Converting the filtered sum into a sum over the interval [1, q/2 + 1) matches the index set, since for integers 1 ≤ h ≤ q/2 is equivalent to 1 ≤ h < q/2 + 1.
uses
Finset.sum_filter,Finset.sum_congr,Finset.mem_filter,Finset.mem_range,Finset.mem_Icohave hmass : wtc_mass q = ∑ h ∈ Finset.Ico 1 (q/2+1), wtc_term q h := by unfold wtc_mass rw [← Finset.sum_filter] apply Finset.sum_congr · ext h simp only [Finset.mem_filter, Finset.mem_range, Finset.mem_Ico] constructor · rintro ⟨hlt, h1, h2⟩; omega · rintro ⟨h1, h2⟩; omega · intro h _; rfl rw [hmass]
The sum over [1, q/2 + 1) decomposes into the head sum over [1, 40) plus the tail sum over [40, q/2 + 1).
Since q ≥ 179, we have q/2 ≥ 89, so 40 ≤ q/2 + 1. The consecutive intervals [1, 40) and [40, q/2 + 1) are disjoint and their union is [1, q/2 + 1), allowing the sum to be split into two disjoint pieces.
uses
Finset.Ico_disjoint_Ico_consecutive,Finset.Ico_union_Ico_eq_Ico,Nat.div_le_div_right,Finset.sum_unionhave hdisj : Disjoint (Finset.Ico 1 40) (Finset.Ico 40 (q/2+1)) := Finset.Ico_disjoint_Ico_consecutive 1 40 (q/2+1) have hsplit : Finset.Ico 1 (q/2+1) = Finset.Ico 1 40 ∪ Finset.Ico 40 (q/2+1) := by rw [Finset.Ico_union_Ico_eq_Ico] · norm_num · have : 179/2 ≤ q/2 := Nat.div_le_div_right hq; omega rw [hsplit, Finset.sum_union hdisj]
The head sum ∑_{h=1}^{39} wtc_term(q, h) is at most 989/50000.
Since q ≥ 179 is odd, we represent q as 179 + 2*k. For each h ∈ [1, 40), 2*h ≤ 78 ≤ 179, so our term iteration lemma bounds wtc_term(q, h) ≤ wtc_term(179, h). Summing this over all h ∈ [1, 40) and applying the baseline numerical bound wild_triple_partition_mass_179_baseline yields the bound 989/50000.
uses
Finset.sum_le_sum,Finset.mem_Ico,wild_triple_partition_mass_le_of_odd_ge_179_term_iter,le_trans,wild_triple_partition_mass_179_baselinehave hhead : ∑ h ∈ Finset.Ico 1 40, wtc_term q h ≤ (989/50000 : ℝ) := by have h1 : ∑ h ∈ Finset.Ico 1 40, wtc_term q h ≤ ∑ h ∈ Finset.Ico 1 40, wtc_term 179 h := by obtain ⟨k, hk⟩ : ∃ k : ℕ, q = 179 + 2*k := by use (q - 179) / 2 omega rw [hk] apply Finset.sum_le_sum intro h hh simp only [Finset.mem_Ico] at hh obtain ⟨h1, h2⟩ := hh exact wild_triple_partition_mass_le_of_odd_ge_179_term_iter h (by omega) (by omega) k exact le_trans h1 wild_triple_partition_mass_179_baseline
The tail sum ∑_{h=40}^{⌊q/2⌋} wtc_term(q, h) is at most 1/10000.
For each h in [40, q/2 + 1), the geometric tail lemma wild_triple_partition_term_geometric_tail bounds wtc_term(q, h) by 3 * h * (27/40)^h. Factoring out 3, the sum ∑_{h=40}^{⌊q/2⌋} h * (27/40)^h is bounded by g(40) using wild_triple_partition_mass_le_of_odd_ge_179_sum_Ico_geom_le with r = 27/40. Direct numerical evaluation with norm_num shows 3 * g(40) ≤ 1/10000.
uses
Nat.div_mul_le_self,Finset.sum_le_sum,Finset.mem_Ico,wild_triple_partition_term_geometric_tail,Nat.div_le_div_right,wild_triple_partition_mass_le_of_odd_ge_179_sum_Ico_geom_le,Finset.mul_sum,Finset.sum_congr,mul_le_mul_of_nonneg_lefthave htail : ∑ h ∈ Finset.Ico 40 (q/2+1), wtc_term q h ≤ (1/10000 : ℝ) := by have h2q : 2*(q/2) ≤ q := by have := Nat.div_mul_le_self q 2 omega have h2 : ∑ h ∈ Finset.Ico 40 (q/2+1), wtc_term q h ≤ ∑ h ∈ Finset.Ico 40 (q/2+1), 3*(h : ℝ)*(27/40 : ℝ)^h := by apply Finset.sum_le_sum intro h hh simp only [Finset.mem_Ico] at hh obtain ⟨h40, hlt⟩ := hh have hle : h ≤ q/2 := by omega exact wild_triple_partition_term_geometric_tail q h (by omega) hodd (by omega) (by omega) have h4 : ∑ h ∈ Finset.Ico 40 (q/2+1), (h : ℝ)*(27/40 : ℝ)^h ≤ ((40:ℝ)/(1-(27/40)) + (27/40)/(1-(27/40))^2) * (27/40)^40 := by have hb : 40 ≤ q/2+1 := by have : 179/2 ≤ q/2 := Nat.div_le_div_right hq; omega exact wild_triple_partition_mass_le_of_odd_ge_179_sum_Ico_geom_le (27/40) (by norm_num) (by norm_num) 40 (q/2+1) hb calc ∑ h ∈ Finset.Ico 40 (q/2+1), wtc_term q h ≤ ∑ h ∈ Finset.Ico 40 (q/2+1), 3*(h : ℝ)*(27/40 : ℝ)^h := h2 _ = 3 * ∑ h ∈ Finset.Ico 40 (q/2+1), (h : ℝ)*(27/40 : ℝ)^h := by rw [Finset.mul_sum] apply Finset.sum_congr rfl; intro h _; ring _ ≤ 3 * (((40:ℝ)/(1-(27/40)) + (27/40)/(1-(27/40))^2) * (27/40)^40) := by apply mul_le_mul_of_nonneg_left h4; norm_num _ ≤ 1/10000 := by norm_num
The total mass wtc_mass(q) is at most 1/50.
Combining the head bound of 989/50000 and the tail bound of 1/10000 gives wtc_mass(q) ≤ 989/50000 + 1/10000 = 994/50000 = 497/25000 < 1/50, which linarith verifies by linear arithmetic.
linarith
Finset.sum_Ico_succ_top- For any function f on integers, the sum over [a, b+1) equals the sum over [a, b) plus the value at b, provided a ≤ b.
pow_succ- For any element a and natural number n, a^(n+1) = a^n * a.
Finset.sum_congr- If two finite sets are equal and two functions agree on every element of the set, their sums over the set are equal.
wild_triple_partition_mass_le_of_odd_ge_179_sum_Ico_telescope_sub- For any sequence f and integers a ≤ b, the telescoping sum ∑_{k=a}^{b-1} (f(k) - f(k+1)) = f(a) - f(b).
div_nonneg- If x ≥ 0 and y ≥ 0, then x / y ≥ 0.
Nat.cast_nonneg- The coercion of any natural number into the real numbers is non-negative.
pow_nonneg- If x ≥ 0, then x^n ≥ 0 for any natural number exponent n.
mul_nonneg- If x ≥ 0 and y ≥ 0, then x * y ≥ 0.
add_nonneg- If x ≥ 0 and y ≥ 0, then x + y ≥ 0.
wild_triple_partition_term_antitone_odd_step- For odd q ≥ 3 and 1 ≤ h ≤ q/2, wtc_term(q+2, h) ≤ wtc_term(q, h).
le_trans- If x ≤ y and y ≤ z, then x ≤ z.
Finset.sum_filter- The sum of f over elements satisfying a predicate equals the sum over the entire set of (if p(x) then f(x) else 0).
Finset.mem_filter- An element belongs to a filtered set if and only if it belongs to the base set and satisfies the filter predicate.
Finset.mem_range- A natural number m belongs to range n if and only if m < n.
Finset.mem_Ico- An integer x belongs to the half-open interval [a, b) if and only if a ≤ x and x < b.
Finset.Ico_disjoint_Ico_consecutive- The consecutive half-open intervals [a, b) and [b, c) are disjoint.
Finset.Ico_union_Ico_eq_Ico- If a ≤ b ≤ c, the union of consecutive half-open intervals [a, b) ∪ [b, c) equals [a, c).
Nat.div_le_div_right- If a ≤ b, then ⌊a/c⌋ ≤ ⌊b/c⌋ for any natural divisor c.
Finset.sum_union- For disjoint finite sets s₁ and s₂, the sum of f over s₁ ∪ s₂ is the sum over s₁ plus the sum over s₂.
Finset.sum_le_sum- If f(i) ≤ g(i) for all i in a finite set s, then ∑_{i ∈ s} f(i) ≤ ∑_{i ∈ s} g(i).
wild_triple_partition_mass_le_of_odd_ge_179_term_iter- For 1 ≤ h ≤ 179/2 and any natural number k, wtc_term(179 + 2*k, h) ≤ wtc_term(179, h).
wild_triple_partition_mass_179_baseline- The head sum ∑_{h=1}^{39} wtc_term(179, h) is at most 989/50000.
Nat.div_mul_le_self- For natural numbers m and n, (m / n) * n ≤ m.
wild_triple_partition_term_geometric_tail- For odd q ≥ 3 and 5 ≤ h ≤ q/2, wtc_term(q, h) ≤ 3*h*(27/40)^h.
wild_triple_partition_mass_le_of_odd_ge_179_sum_Ico_geom_le- For 0 ≤ r < 1 and a ≤ b, the sum ∑_{k=a}^{b-1} k*r^k is at most g(a), where g(n) = (n/(1-r) + r/(1-r)^2) * r^n.
Finset.mul_sum- Multiplication by a scalar distributes over a finite sum: a * ∑_{i ∈ s} f(i) = ∑_{i ∈ s} a * f(i).
mul_le_mul_of_nonneg_left- If b ≤ c and a ≥ 0, then a * b ≤ a * c.
- The proof relies on two externally established properties of wtc_term: the odd-step antitone property (wild_triple_partition_term_antitone_odd_step), which allows bounding the head terms by their values at q = 179, and the geometric tail bound (wild_triple_partition_term_geometric_tail), as well as an explicit numerical calculation at q = 179 (wild_triple_partition_mass_179_baseline).
Written by a model (gemini-3.8-flash) from the Lean proof and the goal states Lean reported at each step. Lean verified the proof; this explanation is a reading aid and is not itself machine-checked.
Lean source view module on GitHub
Choose a tactic — a marked line — to see the goals Lean had before and after it, as in Lean’s infoview. Step with ← → or j k.
open Finsetopen scoped BigOperatorsprivate lemma wild_triple_partition_mass_le_of_odd_ge_179_sum_Ico_telescope_sub (f : ℕ → ℝ) (a b : ℕ) (h : a ≤ b) : ∑ k ∈ Finset.Ico a b, (f k - f (k + 1)) = f a - f b := by induction' h with b' hb' ih · simp · rw [Finset.sum_Ico_succ_top hb', ih] ringprivate lemma wild_triple_partition_mass_le_of_odd_ge_179_sum_Ico_geom_le (r : ℝ) (hr0 : 0 ≤ r) (hr1 : r < 1) (a b : ℕ) (hab : a ≤ b) : let g : ℕ → ℝ := fun n => ((n : ℝ) / (1 - r) + r / (1 - r)^2) * r^n ∑ k ∈ Finset.Ico a b, (k : ℝ) * r^k ≤ g a := by intro g have hdiff : ∀ k : ℕ, (k : ℝ) * r^k = g k - g (k + 1) := by intro k dsimp [g] rw [pow_succ r k] have h1 : ((k + 1 : ℕ) : ℝ) = (k : ℝ) + 1 := by push_cast; rfl rw [h1] have hr : 1 - r ≠ 0 := by linarith have h3 : ((k : ℝ) / (1 - r) + r / (1 - r)^2) - (((k : ℝ) + 1) / (1 - r) + r / (1 - r)^2) * r = (k : ℝ) := by field_simp ring calc (k : ℝ) * r^k = (((k : ℝ) / (1 - r) + r / (1 - r)^2) - (((k : ℝ) + 1) / (1 - r) + r / (1 - r)^2) * r) * r^k := by rw [h3] _ = ((k : ℝ) / (1 - r) + r / (1 - r)^2) * r^k - (((k : ℝ) + 1) / (1 - r) + r / (1 - r)^2) * (r^k * r) := by ring have hcongr : ∑ k ∈ Finset.Ico a b, (k : ℝ) * r^k = ∑ k ∈ Finset.Ico a b, (g k - g (k + 1)) := by apply Finset.sum_congr rfl; intro k _; exact hdiff k rw [hcongr, wild_triple_partition_mass_le_of_odd_ge_179_sum_Ico_telescope_sub g a b hab] have hgb : 0 ≤ g b := by dsimp [g] have hb1 : 0 ≤ (b:ℝ)/(1-r) := div_nonneg (Nat.cast_nonneg b) (by linarith) have hb2 : 0 ≤ r/(1-r)^2 := div_nonneg hr0 (pow_nonneg (by linarith) 2) exact mul_nonneg (add_nonneg hb1 hb2) (pow_nonneg hr0 b) linarithprivate lemma wild_triple_partition_mass_le_of_odd_ge_179_term_iter (h : ℕ) (hh : 1 ≤ h) (hhq : 2*h ≤ 179) : ∀ k : ℕ, wtc_term (179 + 2*k) h ≤ wtc_term 179 h := by intro k induction k with | zero => simp | succ k ih => have hmono : wtc_term (179 + 2*k + 2) h ≤ wtc_term (179 + 2*k) h := wild_triple_partition_term_antitone_odd_step (179 + 2*k) h (by omega) (by omega) hh (by omega) have heq : 179 + 2*(k+1) = 179 + 2*k + 2 := by ring rw [heq] exact le_trans hmono ihtheorem wild_triple_partition_mass_le_of_odd_ge_179 (q : ℕ) (hq : 179 ≤ q) (hodd : q%2 = 1) : wtc_mass q ≤ (1/50 : ℝ) := by have hmass : wtc_mass q = ∑ h ∈ Finset.Ico 1 (q/2+1), wtc_term q h := by unfold wtc_mass rw [← Finset.sum_filter] apply Finset.sum_congr · ext h simp only [Finset.mem_filter, Finset.mem_range, Finset.mem_Ico] constructor · rintro ⟨hlt, h1, h2⟩; omega · rintro ⟨h1, h2⟩; omega · intro h _; rfl rw [hmass] have hdisj : Disjoint (Finset.Ico 1 40) (Finset.Ico 40 (q/2+1)) := Finset.Ico_disjoint_Ico_consecutive 1 40 (q/2+1) have hsplit : Finset.Ico 1 (q/2+1) = Finset.Ico 1 40 ∪ Finset.Ico 40 (q/2+1) := by rw [Finset.Ico_union_Ico_eq_Ico] · norm_num · have : 179/2 ≤ q/2 := Nat.div_le_div_right hq; omega rw [hsplit, Finset.sum_union hdisj] have hhead : ∑ h ∈ Finset.Ico 1 40, wtc_term q h ≤ (989/50000 : ℝ) := by have h1 : ∑ h ∈ Finset.Ico 1 40, wtc_term q h ≤ ∑ h ∈ Finset.Ico 1 40, wtc_term 179 h := by obtain ⟨k, hk⟩ : ∃ k : ℕ, q = 179 + 2*k := by use (q - 179) / 2 omega rw [hk] apply Finset.sum_le_sum intro h hh simp only [Finset.mem_Ico] at hh obtain ⟨h1, h2⟩ := hh exact wild_triple_partition_mass_le_of_odd_ge_179_term_iter h (by omega) (by omega) k exact le_trans h1 wild_triple_partition_mass_179_baseline have htail : ∑ h ∈ Finset.Ico 40 (q/2+1), wtc_term q h ≤ (1/10000 : ℝ) := by have h2q : 2*(q/2) ≤ q := by have := Nat.div_mul_le_self q 2 omega have h2 : ∑ h ∈ Finset.Ico 40 (q/2+1), wtc_term q h ≤ ∑ h ∈ Finset.Ico 40 (q/2+1), 3*(h : ℝ)*(27/40 : ℝ)^h := by apply Finset.sum_le_sum intro h hh simp only [Finset.mem_Ico] at hh obtain ⟨h40, hlt⟩ := hh have hle : h ≤ q/2 := by omega exact wild_triple_partition_term_geometric_tail q h (by omega) hodd (by omega) (by omega) have h4 : ∑ h ∈ Finset.Ico 40 (q/2+1), (h : ℝ)*(27/40 : ℝ)^h ≤ ((40:ℝ)/(1-(27/40)) + (27/40)/(1-(27/40))^2) * (27/40)^40 := by have hb : 40 ≤ q/2+1 := by have : 179/2 ≤ q/2 := Nat.div_le_div_right hq; omega exact wild_triple_partition_mass_le_of_odd_ge_179_sum_Ico_geom_le (27/40) (by norm_num) (by norm_num) 40 (q/2+1) hb calc ∑ h ∈ Finset.Ico 40 (q/2+1), wtc_term q h ≤ ∑ h ∈ Finset.Ico 40 (q/2+1), 3*(h : ℝ)*(27/40 : ℝ)^h := h2 _ = 3 * ∑ h ∈ Finset.Ico 40 (q/2+1), (h : ℝ)*(27/40 : ℝ)^h := by rw [Finset.mul_sum] apply Finset.sum_congr rfl; intro h _; ring _ ≤ 3 * (((40:ℝ)/(1-(27/40)) + (27/40)/(1-(27/40))^2) * (27/40)^40) := by apply mul_le_mul_of_nonneg_left h4; norm_num _ ≤ 1/10000 := by norm_num linarith