Description: In the recursive scheme
"(n+1)-xor" <-> if- ( ph , -. "n-xor" , "n-xor" )
we set n = 0 to formally arrive at an expression for "1-xor". The base case "0-xor" is replaced with F. , as a sequence of 0 inputs never has an odd number being part of it. (Contributed by Wolf Lammen, 11-May-2024)
Ref | Expression | ||
---|---|---|---|
Assertion | wl-1xor | ⊢ ( if- ( 𝜓 , ¬ ⊥ , ⊥ ) ↔ 𝜓 ) |
Step | Hyp | Ref | Expression |
---|---|---|---|
1 | tbtru | ⊢ ( 𝜓 ↔ ( 𝜓 ↔ ⊤ ) ) | |
2 | 1 | biimpi | ⊢ ( 𝜓 → ( 𝜓 ↔ ⊤ ) ) |
3 | notfal | ⊢ ( ¬ ⊥ ↔ ⊤ ) | |
4 | 2 3 | bitr4di | ⊢ ( 𝜓 → ( 𝜓 ↔ ¬ ⊥ ) ) |
5 | nbfal | ⊢ ( ¬ 𝜓 ↔ ( 𝜓 ↔ ⊥ ) ) | |
6 | 5 | biimpi | ⊢ ( ¬ 𝜓 → ( 𝜓 ↔ ⊥ ) ) |
7 | 4 6 | casesifp | ⊢ ( 𝜓 ↔ if- ( 𝜓 , ¬ ⊥ , ⊥ ) ) |
8 | 7 | bicomi | ⊢ ( if- ( 𝜓 , ¬ ⊥ , ⊥ ) ↔ 𝜓 ) |