ProjectIdeas

Project Ideas

Now that you know the basics of Coq you are itching to prove something. Here are some ideas.

Other than formalizing existing proofs, one useful exercise is to update some of Coq's user-contributed libraries so that they build on the recent additions in the Coq StandardLibrary rather than their own foundations.

See Top100MathematicalTheoremsInCoq and Formalizing 100 Theorems

Fermat Last Theorem for n = 3

Statement

forall (x y z:Z), x^3 + y^3 = z^3 -> x=0 \/ y= 0 \/ z=0

Resources

See Fermat's Last Theorem: Proof for n=3.

http://nshmyrev.narod.ru/temp/fermat4.tar.gz start of the proof.

Estimated Difficulty

|{}| |{}| |{o}| |{o}| |{o}| |{o}| |{o}| |{o}| (change the estimate if you disagree)

Bounties Offered

(none)

Fermat Last Theorem

Statement

forall (x y z:Z) (n:nat), x^(n+3) + y^(n+3) = z^(n+3) -> x=0 \/ y= 0 \/ z=0

Resources

Fermat's last theorem has been proven in Coq for the n=4 case. The proof is available in UserContributions/CNAM/Fermat4/ .

See also, Computer verification of Wiles' proof of Fermat's Last Theorem

Estimated Difficulty

|{}| |{}| |{}| |{}| |{}| |{}| |{}| |{}| (change the estimate if you disagree)

Bounties Offered

(none)

Lusin Separation Theorem

Informal Statement

Any strongly disjoint pair of analytic sets is strongly Borel sepearable.

Comments

Peter Aczel has a proof in constructive ZF. It should be possible to translate this proof into type theory.

Resources

• http://foundations.cs.ru.nl/fndswiki/A_constructive_version_of_the_Lusin_Separation_Theorem

Estimated Difficulty

|{}| |{}| |{}| |{}| |{o}| |{o}| |{o}| |{o}| (change the estimate if you disagree)

Conway's Cosmological Theorem

Informal Statement

The maximal length of an atom in the splitting of a 50-day-old string is 74. Every such atom decays, in at most 17 days, into stable or transuranic elements.

Resources

http://www.cs.cmu.edu/~kw/pubs/conway.pdf http://www.mathematik.uni-bielefeld.de/~sillke/SEQUENCES/series000

Estimated Difficulty

??? (change the estimate if you disagree)

Bounties Offered

(none)

Completed Projects

Tarski's Undefinability of Truth

This has now been completed by Wouter Stekelenburg.

Elliptic Curve Primality Proving

This has now been completed. See http://coqprime.gforge.inria.fr/