Zulip Chat Archive
Stream: new members
Topic: Matching on Fin
Dax Fohl (Jul 22 2023 at 16:58):
I've been banging my head against the wall, but can't get anything like this to work. It works fine if I change k
to Nat
. But I can't get it to work for Fin
.
def bounded_decrement (k: Fin 20) : Nat :=
match k with
| Nat.zero => 0
| Nat.succ n => n
Dax Fohl (Jul 22 2023 at 17:01):
I'm ultimately looking for def bounded_decrement (n: Nat) (k: Fin n) : Fin (n - 1)
Yury G. Kudryashov (Jul 22 2023 at 17:17):
Niels Voss (Jul 22 2023 at 19:51):
For the first half of your question, maybe something like this?
def bounded_decrement (k: Fin 20) : Nat :=
match k with
| ⟨Nat.zero, hk⟩ => 0
| ⟨Nat.succ n, hk⟩ => n
hk
will be a proof that the natural number is less than 20
. What you have to be put in the match
clause must be the same as how Fin
is defined. A term k
of type Fin n
is essentially a Nat
and a proof that k < n
, so to match on it you have to have a variable that stores that proof.
Niels Voss (Jul 22 2023 at 19:54):
(In case it isn't clear, ⟨Nat.zero, hk⟩
is in this context shorthand for Fin.mk (Nat.zero) hk
and ⟨Nat.succ n, hk⟩
is shorthand for Fin.mk (Nat.succ n) hk
.)
Dax Fohl (Jul 23 2023 at 02:16):
Thanks! yeah I stumbled across that notation when looking at the def for Fin.succ
.
Niels Voss (Jul 23 2023 at 04:27):
You could also check out Fin2
, which is like Fin
but defined recursively. I don't know when you'd use it, but I found it interesting
Last updated: Dec 20 2023 at 11:08 UTC