Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

The place attached to a noncuspidal rational canonical point

Definition
N25_CanonicalNoncuspPlace

by FormallyCorrect · Nov 16, 2026 · Mathlib 0df444a (Lean v4.33.1)

mazur-n25

This file packages the coordinate and denominator conditions that distinguish the five rational cusps from the dense birational open. For a rational canonical point [x:y:z:w][x:y:z:w][x:y:z:w] outside those cusp classes, it defines the associated place by the plane coordinates

u=x/(y+z−w),v=x2/(z(y+z−w))u=x/(y+z-w), v=x^2/(z(y+z-w))u=x/(y+z−w),v=x2/(z(y+z−w))

and transports the resulting plane place through the canonical function-field equivalence. This converts a hypothetical noncuspidal rational point into the place used by the modular endgame.

Definition code
import Mathlib
import Definitions.Def_RationalPointsN25CanonicalPoints
import Definitions.Def_N25_CanonicalUnitPlane
import Definitions.Def_N25_UnitPlanePointPlace
/-! Rational noncuspidal canonical points define places through the explicit birational map.

A rational point of the canonical model outside the five cusp classes has all coordinates nonzero and nonzero
denominators of the birational map; these facts are proved here because the place is defined through that map. -/

/-! ### Boundary points of the canonical model are cusps -/
noncomputable section

/-! Proofs of the lemmas stated in `Theorems.Thm_CanonicalLevelTwentyFive_BoundaryPoints` and their helpers. -/

namespace MazurProof.RationalPointsN25CanonicalPoints

open Polynomial

/-- The canonical cubic is linear in `y` with the displayed coefficient and
constant term. -/
theorem canonicalCubic25_eq_linear (x y z w : ℚ) :
    canonicalCubic25 x y z w =
      y * projectionDenominator25 x z w + projectionNumerator25 x z w := by
  simp only [canonicalCubic25, projectionDenominator25, projectionNumerator25]
  ring

/-- Exact elimination identity behind the plane sextic projection. -/
theorem canonical_elimination_identity25 (x y z w : ℚ) :
    canonicalPlaneSextic25 x z w =
      projectionDenominator25 x z w ^ 2 * canonicalQuadric25 x y z w +
      (projectionNumerator25 x z w -
          projectionDenominator25 x z w * y -
          z * projectionDenominator25 x z w) *
        canonicalCubic25 x y z w := by
  simp only [canonicalPlaneSextic25, projectionDenominator25,
    projectionNumerator25, canonicalQuadric25, canonicalCubic25]
  ring

/-- Every point on the canonical complete intersection projects to the plane
sextic. -/
theorem canonical_point_projects_to_plane_sextic {x y z w : ℚ}
    (hQ : canonicalQuadric25 x y z w = 0)
    (hC : canonicalCubic25 x y z w = 0) :
    canonicalPlaneSextic25 x z w = 0 := by
  rw [canonical_elimination_identity25, hQ, hC]
  ring

/-- Conversely, away from the linear-projection denominator, every rational
point on the plane sextic lifts to the canonical complete intersection. -/
theorem plane_sextic_lifts_to_canonical {x z w : ℚ}
    (hD : projectionDenominator25 x z w ≠ 0)
    (hF : canonicalPlaneSextic25 x z w = 0) :
    canonicalQuadric25 x (recoveredY25 x z w) z w = 0 ∧
      canonicalCubic25 x (recoveredY25 x z w) z w = 0 := by
  have hC : canonicalCubic25 x (recoveredY25 x z w) z w = 0 := by
    rw [canonicalCubic25_eq_linear]
    unfold recoveredY25
    field_simp [hD]
    ring
  refine ⟨?_, hC⟩
  have hid := canonical_elimination_identity25 x (recoveredY25 x z w) z w
  rw [hF, hC, mul_zero, add_zero] at hid
  exact (mul_eq_zero.mp hid.symm).resolve_left (pow_ne_zero 2 hD)

/-- The eliminated coordinate is uniquely recovered from any canonical point
on the nonexceptional projection locus. -/
theorem recoveredY25_eq_of_canonical {x y z w : ℚ}
    (hD : projectionDenominator25 x z w ≠ 0)
    (hC : canonicalCubic25 x y z w = 0) :
    y = recoveredY25 x z w := by
  rw [canonicalCubic25_eq_linear] at hC
  unfold recoveredY25
  field_simp [hD]
  linarith

/-! ## Elementary projective-coordinate helpers -/

private theorem cusp0001_of_coordinates {x y z w : ℚ}
    (hnz : NonzeroQuadruple x y z w)
    (hx : x = 0) (hy : y = 0) (hz : z = 0) :
    IsCanonicalCusp25 x y z w := by
  subst x
  subst y
  subst z
  have hw : w ≠ 0 := by simpa [NonzeroQuadruple] using hnz
  exact Or.inl ⟨w, hw, by norm_num, by norm_num, by norm_num, by ring⟩

private theorem cusp0m110_of_coordinates {x y z w : ℚ}
    (hnz : NonzeroQuadruple x y z w)
    (hx : x = 0) (hyz : y = -z) (hw : w = 0) :
    IsCanonicalCusp25 x y z w := by
  subst x
  subst y
  subst w
  have hz : z ≠ 0 := by simpa [NonzeroQuadruple] using hnz
  exact Or.inr <| Or.inl ⟨z, hz, by norm_num, by ring, by ring, by norm_num⟩

private theorem cusp1101_of_coordinates {x y z w : ℚ}
    (hnz : NonzeroQuadruple x y z w)
    (hy : y = x) (hz : z = 0) (hw : w = x) :
    IsCanonicalCusp25 x y z w := by
  subst y
  subst z
  subst w
  have hx : x ≠ 0 := by simpa [NonzeroQuadruple] using hnz
  exact Or.inr <| Or.inr <| Or.inl
    ⟨x, hx, by ring, by ring, by norm_num, by ring⟩

private theorem cusp0010_of_coordinates {x y z w : ℚ}
    (hnz : NonzeroQuadruple x y z w)
    (hx : x = 0) (hy : y = 0) (hw : w = 0) :
    IsCanonicalCusp25 x y z w := by
  subst x
  subst y
  subst w
  have hz : z ≠ 0 := by simpa [NonzeroQuadruple] using hnz
  exact Or.inr <| Or.inr <| Or.inr <| Or.inl
    ⟨z, hz, by norm_num, by norm_num, by ring, by norm_num⟩

private theorem cusp1000_of_coordinates {x y z w : ℚ}
    (hnz : NonzeroQuadruple x y z w)
    (hy : y = 0) (hz : z = 0) (hw : w = 0) :
    IsCanonicalCusp25 x y z w := by
  subst y
  subst z
  subst w
  have hx : x ≠ 0 := by simpa [NonzeroQuadruple] using hnz
  exact Or.inr <| Or.inr <| Or.inr <| Or.inr
    ⟨x, hx, by ring, by norm_num, by norm_num, by norm_num⟩

/-! ## The cubic obstruction in the `y = 0` chart -/

/-- The monic cubic `T^3 - T - 1` has no rational root. -/
theorem plastic_cubic_ne_zero (t : ℚ) : t ^ 3 - t - 1 ≠ 0 := by
  intro ht
  let p : ℤ[X] := X ^ 3 - X - 1
  have hp : p.Monic := by
    dsimp [p]
    monicity
    all_goals norm_num
  have heval : Polynomial.aeval t p = 0 := by
    norm_num [p, Polynomial.aeval_def]
    exact ht
  obtain ⟨n, hn, hndvd⟩ :=
    exists_integer_of_is_root_of_monic hp heval
  have hunit : IsUnit n := by
    apply isUnit_of_dvd_one
    have hneg : n ∣ (-1 : ℤ) := by simpa [p] using hndvd
    simpa using hneg.neg_right
  rcases Int.isUnit_iff.mp hunit with hnone | hnneg
  · have htone : t = 1 := by simpa [hnone] using hn
    rw [htone] at ht
    norm_num at ht
  · have htneg : t = -1 := by simpa [hnneg] using hn
    rw [htneg] at ht
    norm_num at ht

/-- A second monic cubic occurring in the exceptional projection locus has no
rational root. -/
theorem projection_cubic_ne_zero (t : ℚ) :
    t ^ 3 - 2 * t ^ 2 + 3 * t - 1 ≠ 0 := by
  intro ht
  let p : ℤ[X] := X ^ 3 - 2 * X ^ 2 + 3 * X - 1
  have hp : p.Monic := by
    dsimp [p]
    monicity
    all_goals norm_num
  have heval : Polynomial.aeval t p = 0 := by
    norm_num [p, Polynomial.aeval_def]
    exact ht
  obtain ⟨n, hn, hndvd⟩ :=
    exists_integer_of_is_root_of_monic hp heval
  have hunit : IsUnit n := by
    apply isUnit_of_dvd_one
    have hneg : n ∣ (-1 : ℤ) := by simpa [p] using hndvd
    simpa using hneg.neg_right
  rcases Int.isUnit_iff.mp hunit with hnone | hnneg
  · have htone : t = 1 := by simpa [hnone] using hn
    rw [htone] at ht
    norm_num at ht
  · have htneg : t = -1 := by simpa [hnneg] using hn
    rw [htneg] at ht
    norm_num at ht

/-! ## Classification on the four coordinate hyperplanes -/

/-- Every rational canonical point on the hyperplane `w = 0` is a cusp. -/
theorem canonical_point_w_zero_is_cusp {x y z w : ℚ}
    (hnz : NonzeroQuadruple x y z w)
    (hQ : canonicalQuadric25 x y z w = 0)
    (hC : canonicalCubic25 x y z w = 0)
    (hw : w = 0) :
    IsCanonicalCusp25 x y z w := by
  subst w
  have hxyz : x * y * z = 0 := by
    simpa [canonicalCubic25] using hC
  rcases mul_eq_zero.mp hxyz with hxy | hz
  · rcases mul_eq_zero.mp hxy with hx | hy
    · have hyfac : y * (y + z) = 0 := by
        simp only [canonicalQuadric25, hx, zero_mul, neg_zero] at hQ
        nlinarith
      rcases mul_eq_zero.mp hyfac with hy | hyz
      · exact cusp0010_of_coordinates hnz hx hy rfl
      · have hyneg : y = -z := by linarith
        exact cusp0m110_of_coordinates hnz hx hyneg rfl
    · have hxz : x * z = 0 := by
        simp only [canonicalQuadric25, hy, zero_mul, add_zero] at hQ
        nlinarith
      rcases mul_eq_zero.mp hxz with hx | hz
      · exact cusp0010_of_coordinates hnz hx hy rfl
      · exact cusp1000_of_coordinates hnz hy hz rfl
  · have hy : y = 0 := by
      simp only [canonicalQuadric25, hz, mul_zero, add_zero] at hQ
      nlinarith [sq_nonneg y]
    exact cusp1000_of_coordinates hnz hy hz rfl

/-- Every rational canonical point on the hyperplane `z = 0` is a cusp. -/
theorem canonical_point_z_zero_is_cusp {x y z w : ℚ}
    (hnz : NonzeroQuadruple x y z w)
    (hQ : canonicalQuadric25 x y z w = 0)
    (hC : canonicalCubic25 x y z w = 0)
    (hz : z = 0) :
    IsCanonicalCusp25 x y z w := by
  subst z
  have hC' : x ^ 2 * w - x * y * w = 0 := by
    simpa [canonicalCubic25] using hC
  have hprod : x * w * (x - y) = 0 := by
    linear_combination hC'
  rcases mul_eq_zero.mp hprod with hxw | hxy
  · rcases mul_eq_zero.mp hxw with hx | hw
    · have hy : y = 0 := by
        simp only [canonicalQuadric25, hx, zero_mul, neg_zero] at hQ
        nlinarith [sq_nonneg y]
      exact cusp0001_of_coordinates hnz hx hy rfl
    · have hy : y = 0 := by
        simp only [canonicalQuadric25, hw, mul_zero, add_zero] at hQ
        nlinarith [sq_nonneg y]
      exact cusp1000_of_coordinates hnz hy rfl hw
  · have hy : y = x := by linarith
    have hxw : x * (x - w) = 0 := by
      simp only [canonicalQuadric25, hy] at hQ
      nlinarith
    rcases mul_eq_zero.mp hxw with hx | hxw
    · have hy0 : y = 0 := by rw [hy, hx]
      exact cusp0001_of_coordinates hnz hx hy0 rfl
    · have hw : w = x := by linarith
      exact cusp1101_of_coordinates hnz hy rfl hw

/-- Every rational canonical point on the hyperplane `x = 0` is a cusp. -/
theorem canonical_point_x_zero_is_cusp {x y z w : ℚ}
    (hnz : NonzeroQuadruple x y z w)
    (hQ : canonicalQuadric25 x y z w = 0)
    (hC : canonicalCubic25 x y z w = 0)
    (hx : x = 0) :
    IsCanonicalCusp25 x y z w := by
  subst x
  have hC' : y * z * w + z ^ 2 * w - z * w ^ 2 = 0 := by
    simpa [canonicalCubic25] using hC
  have hprod : z * w * (y + z - w) = 0 := by
    linear_combination hC'
  rcases mul_eq_zero.mp hprod with hzw | hyzw
  · rcases mul_eq_zero.mp hzw with hz | hw
    · have hy : y = 0 := by
        simp only [canonicalQuadric25, hz, mul_zero, add_zero] at hQ
        nlinarith [sq_nonneg y]
      exact cusp0001_of_coordinates hnz rfl hy hz
    · have hyfac : y * (y + z) = 0 := by
        simp only [canonicalQuadric25, hw, mul_zero, add_zero] at hQ
        nlinarith
      rcases mul_eq_zero.mp hyfac with hy | hyz
      · exact cusp0010_of_coordinates hnz rfl hy hw
      · have hyneg : y = -z := by linarith
        exact cusp0m110_of_coordinates hnz rfl hyneg hw
  · have hwzero : w = 0 := by
      have hyzw' : y + z = w := by linarith
      have hQ' : y ^ 2 + y * z + z * w = 0 := by
        simpa [canonicalQuadric25] using hQ
      rw [← hyzw'] at hQ'
      have hsum : y + z = 0 := by nlinarith [sq_nonneg (y + z)]
      linarith
    have hyfac : y * (y + z) = 0 := by
      simp only [canonicalQuadric25, hwzero, mul_zero, add_zero] at hQ
      nlinarith
    rcases mul_eq_zero.mp hyfac with hy | hyz
    · exact cusp0010_of_coordinates hnz rfl hy hwzero
    · have hyneg : y = -z := by linarith
      exact cusp0m110_of_coordinates hnz rfl hyneg hwzero

/-- Every rational canonical point on the hyperplane `y = 0` is a cusp.

After the easy component `w = 0`, elimination of `x` from the two equations
gives

`z * (z^3 - w^2*z - w^3) = 0`.

The second factor would make `z/w` a rational root of `T^3 - T - 1`. -/
theorem canonical_point_y_zero_is_cusp {x y z w : ℚ}
    (hnz : NonzeroQuadruple x y z w)
    (hQ : canonicalQuadric25 x y z w = 0)
    (hC : canonicalCubic25 x y z w = 0)
    (hy : y = 0) :
    IsCanonicalCusp25 x y z w := by
  subst y
  by_cases hw : w = 0
  · exact canonical_point_w_zero_is_cusp hnz hQ hC hw
  have hQ' : -x * z - x * w + z * w = 0 := by
    simpa [canonicalQuadric25] using hQ
  have hC' : x ^ 2 * w - x * z * w + z ^ 2 * w - z * w ^ 2 = 0 := by
    simpa [canonicalCubic25] using hC
  have hinner : x ^ 2 - x * z + z ^ 2 - z * w = 0 := by
    have hfactor : w * (x ^ 2 - x * z + z ^ 2 - z * w) = 0 := by
      linear_combination hC'
    exact (mul_eq_zero.mp hfactor).resolve_left hw
  have hres : z * (z ^ 3 - w ^ 2 * z - w ^ 3) = 0 := by
    linear_combination
      ((w + z) * x - z ^ 2) * hQ' + (w + z) ^ 2 * hinner
  rcases mul_eq_zero.mp hres with hz | hhom
  · have hx : x = 0 := by
      rw [hz] at hQ'
      norm_num at hQ'
      exact hQ'.resolve_right hw
    exact cusp0001_of_coordinates hnz hx rfl hz
  · have hroot : (z / w) ^ 3 - z / w - 1 = 0 := by
      field_simp [hw]
      nlinarith
    exact False.elim ((plastic_cubic_ne_zero (z / w)) hroot)

/-- A rational point on the canonical model with a zero homogeneous
coordinate is one of the five rational cusps. -/
theorem canonical_point_with_zero_coordinate_is_cusp {x y z w : ℚ}
    (hnz : NonzeroQuadruple x y z w)
    (hQ : canonicalQuadric25 x y z w = 0)
    (hC : canonicalCubic25 x y z w = 0)
    (hzero : x = 0 ∨ y = 0 ∨ z = 0 ∨ w = 0) :
    IsCanonicalCusp25 x y z w := by
  rcases hzero with hx | hy | hz | hw
  · exact canonical_point_x_zero_is_cusp hnz hQ hC hx
  · exact canonical_point_y_zero_is_cusp hnz hQ hC hy
  · exact canonical_point_z_zero_is_cusp hnz hQ hC hz
  · exact canonical_point_w_zero_is_cusp hnz hQ hC hw

/-- Contrapositive form used by later affine/descent reductions: every
noncuspidal rational point has all four coordinates nonzero. -/
theorem all_coordinates_ne_zero_of_not_cusp {x y z w : ℚ}
    (hnz : NonzeroQuadruple x y z w)
    (hQ : canonicalQuadric25 x y z w = 0)
    (hC : canonicalCubic25 x y z w = 0)
    (hncusp : ¬ IsCanonicalCusp25 x y z w) :
    x ≠ 0 ∧ y ≠ 0 ∧ z ≠ 0 ∧ w ≠ 0 := by
  constructor
  · intro hx
    exact hncusp (canonical_point_with_zero_coordinate_is_cusp hnz hQ hC (Or.inl hx))
  constructor
  · intro hy
    exact hncusp (canonical_point_with_zero_coordinate_is_cusp hnz hQ hC (Or.inr (Or.inl hy)))
  constructor
  · intro hz
    exact hncusp
      (canonical_point_with_zero_coordinate_is_cusp hnz hQ hC (Or.inr (Or.inr (Or.inl hz))))
  · intro hw
    exact hncusp
      (canonical_point_with_zero_coordinate_is_cusp hnz hQ hC (Or.inr (Or.inr (Or.inr hw))))

/-- The plane-sextic projection denominator cannot vanish at a noncuspidal
rational point.  Indeed, simultaneous vanishing of the denominator and the
canonical cubic eliminates `x` to

`z * (z^3 - 2*w*z^2 + 3*w^2*z - w^3) = 0`.

On the dense torus the first factor is nonzero, while the second factor would
give a rational root of `T^3 - 2*T^2 + 3*T - 1`. -/
theorem projectionDenominator25_ne_zero_of_not_cusp {x y z w : ℚ}
    (hnz : NonzeroQuadruple x y z w)
    (hQ : canonicalQuadric25 x y z w = 0)
    (hC : canonicalCubic25 x y z w = 0)
    (hncusp : ¬ IsCanonicalCusp25 x y z w) :
    projectionDenominator25 x z w ≠ 0 := by
  obtain ⟨_hx, _hy, hz, hw⟩ :=
    all_coordinates_ne_zero_of_not_cusp hnz hQ hC hncusp
  intro hD
  have hD' : x * z - x * w + z * w = 0 := by
    simpa [projectionDenominator25] using hD
  have hdecomp :
      y * projectionDenominator25 x z w +
        w * (x ^ 2 - x * z + z ^ 2 - z * w) = 0 := by
    change y * projectionDenominator25 x z w +
      projectionNumerator25 x z w = 0
    rw [← canonicalCubic25_eq_linear]
    exact hC
  have hwi : w * (x ^ 2 - x * z + z ^ 2 - z * w) = 0 := by
    simpa only [hD, mul_zero, zero_add] using hdecomp
  have hinner : x ^ 2 - x * z + z ^ 2 - z * w = 0 :=
    (mul_eq_zero.mp hwi).resolve_left hw
  have hres :
      z * (z ^ 3 - 2 * w * z ^ 2 + 3 * w ^ 2 * z - w ^ 3) = 0 := by
    linear_combination
      ((w - z) * x + z ^ 2) * hD' + (w - z) ^ 2 * hinner
  have hhom : z ^ 3 - 2 * w * z ^ 2 + 3 * w ^ 2 * z - w ^ 3 = 0 :=
    (mul_eq_zero.mp hres).resolve_left hz
  have hroot :
      (z / w) ^ 3 - 2 * (z / w) ^ 2 + 3 * (z / w) - 1 = 0 := by
    field_simp [hw]
    nlinarith
  exact (projection_cubic_ne_zero (z / w)) hroot

/-- Dense-open birational reduction: every hypothetical noncuspidal canonical
point projects to the plane sextic on its nonexceptional locus, and its fourth
coordinate `y` is recovered by the explicit rational formula. -/
theorem noncuspidal_point_to_plane_sextic {x y z w : ℚ}
    (hnz : NonzeroQuadruple x y z w)
    (hQ : canonicalQuadric25 x y z w = 0)
    (hC : canonicalCubic25 x y z w = 0)
    (hncusp : ¬ IsCanonicalCusp25 x y z w) :
    canonicalPlaneSextic25 x z w = 0 ∧
      projectionDenominator25 x z w ≠ 0 ∧
      y = recoveredY25 x z w := by
  have hD := projectionDenominator25_ne_zero_of_not_cusp hnz hQ hC hncusp
  exact ⟨canonical_point_projects_to_plane_sextic hQ hC, hD,
    recoveredY25_eq_of_canonical hD hC⟩

end MazurProof.RationalPointsN25CanonicalPoints

end

/-! ### Denominators of the birational map at noncuspidal points -/
noncomputable section

/-! Noncuspidal rational canonical points lie in the open used by the birational formulas. -/
namespace N25
open Polynomial

/-- The first birational denominator cannot vanish when x and z are nonzero. -/
theorem canonicalDenominator_ne_zero {k : Type*} [Field k] {x y z w : k}
    (hx : x ≠ 0) (hz : z ≠ 0)
    (hq : canonicalQuadric x y z w = 0) (hc : canonicalCubic x y z w = 0) :
    canonicalDenominator y z w ≠ 0 := by
  intro ha
  unfold canonicalDenominator at ha
  have hw : w = y+z := by linear_combination -ha
  rw [hw] at hq hc
  unfold canonicalQuadric at hq
  unfold canonicalCubic at hc
  have hxy : x*z*(x-y) = 0 := by linear_combination -x*hq-hc
  have hxy' : x = y := sub_eq_zero.mp ((mul_eq_zero.mp hxy).resolve_left (mul_ne_zero hx hz))
  subst y
  have hz2 : z^2 = 0 := by linear_combination hq
  exact (pow_ne_zero 2 hz) hz2

/-- The reverse-composition multiplier is nonzero at rational points with x,z nonzero. -/
theorem canonicalMultiplier_ne_zero {x y z w : ℚ}
    (hx : x ≠ 0) (hz : z ≠ 0)
    (hq : canonicalQuadric x y z w = 0) (hc : canonicalCubic x y z w = 0) :
    canonicalMultiplier x y z w ≠ 0 := by
  have ha := canonicalDenominator_ne_zero hx hz hq hc
  intro hb
  let a := canonicalDenominator y z w
  have hid : z*(x^2-z*a+a^2) - y*canonicalMultiplier x y z w =
      -x*canonicalQuadric x y z w-canonicalCubic x y z w := by
    dsimp [a, canonicalMultiplier, canonicalDenominator, canonicalQuadric, canonicalCubic]
    ring
  rw [hb, hq, hc] at hid
  have hpoly : x^2-z*a+a^2 = 0 := by
    have hm : z*(x^2-z*a+a^2) = 0 := by simpa using hid
    exact (mul_eq_zero.mp hm).resolve_left hz
  have hb' : x*z+z*a+x*a = 0 := hb
  have hcum : x^3+x^2*a+2*x*a^2+a^3 = 0 := by
    linear_combination (x+a)*hpoly+a*hb'
  apply planeCubicBranch_ne_zero (x/a)
  have ha' : a ≠ 0 := ha
  field_simp [ha']
  linear_combination hcum

/-- The September noncusp classification supplies every denominator of the canonical open. -/
theorem canonical_noncuspidal_denominators {x y z w : ℚ}
    (hnz : MazurProof.RationalPointsN25CanonicalPoints.NonzeroQuadruple x y z w)
    (hq : canonicalQuadric x y z w = 0) (hc : canonicalCubic x y z w = 0)
    (hnc : ¬ MazurProof.RationalPointsN25CanonicalPoints.IsCanonicalCusp25 x y z w) :
    x ≠ 0 ∧ z ≠ 0 ∧ canonicalDenominator y z w ≠ 0 ∧ canonicalMultiplier x y z w ≠ 0 := by
  have h := MazurProof.RationalPointsN25CanonicalPoints.all_coordinates_ne_zero_of_not_cusp
    hnz hq hc hnc
  exact ⟨h.1, h.2.2.1, canonicalDenominator_ne_zero h.1 h.2.2.1 hq hc,
    canonicalMultiplier_ne_zero h.1 h.2.2.1 hq hc⟩

end N25

end

/-! ### The place -/
noncomputable section
namespace N25

/-- The place of a rational canonical point outside the five cusp classes. -/
def canonicalNoncuspPlace25 (x y z w : ℚ)
    (hnz : MazurProof.RationalPointsN25CanonicalPoints.NonzeroQuadruple x y z w)
    (hq : canonicalQuadric x y z w = 0) (hc : canonicalCubic x y z w = 0)
    (hnc : ¬ MazurProof.RationalPointsN25CanonicalPoints.IsCanonicalCusp25 x y z w) :
    AlgebraicCurve.Place ℚ (CanonicalFF25 ℚ) :=
  let h := canonical_noncuspidal_denominators hnz hq hc hnc
  canonicalPointPlaceOfUnitPlane25 (x/canonicalDenominator y z w)
    (x^2/(z*canonicalDenominator y z w))
    (canonical_maps_to_plane x y z w h.1 h.2.2.1 h.2.1 hq hc)
    (div_ne_zero h.1 h.2.2.1)

end N25
Source
B. Mazur, Modular curves and the Eisenstein ideal, Publ. IHES 47 (1977); Lean formalization of the order-25 case for this mission. Modules in the MazurProof/Keystone/Tate families are adapted from anthropics/fermats-last-theorem (Apache-2.0) via the xiangyazi24/FLT fork.

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me