First, we define a function f:X→X with
f(3i−2)if(j)=3i−1,f(3i−1)=3i,f(3i)=3i−2,=1,2,…,30,=100,91≤j≤99,f(100)=99.
Obviously, f satisfies condition (1). For any A⊆X with ∣A∣=40, if
(i) there exists an integer i with 1≤i≤30 such that ∣A∩{3i−2,3i−1,3i}∣≥2, then A∩f(A)=∅; or
(ii) 91,92,…,100∈A, then A∩f(A)=∅ also holds.
In both cases, f satisfies condition (2). If a subset B of X satisfies f(B)∪B=X, then we have ∣B∩{3i−2,3i−1,3i}∣≥2 for all 1≤i≤30, {91,92,…,98}⊆B, and B∩{99,100}=∅. Hence, ∣B∣≥69.
Next, we will show that, for any function f satisfying the described conditions, there exists a subset B⊆X with ∣B∣≤69 such that f(B)∪B=X.
Among all the subsets U⊆X with U∩f(U)=∅, choose one such that ∣U∣ is maximal. If there are many U⊆X with ∣U∣ being maximal, choose one such that ∣f(U)∣ is maximal. The existence of U is guaranteed by condition (1). Let V=f(U), W=X∖(U∪V). Note that U,V,W are pairwise disjoint and X=U∪V∪W. From condition (2), ∣U∣≤39, ∣V∣≤39, ∣W∣≥22.
We make the following assertions:
(i) f(w)∈U, for all w∈W. Otherwise, let U′=U∪{w}, since f(U)=V, f(w)∈/U, f(w)=w, we have U′∩f(U′)=∅. It is a contradiction to ∣U∣ being maximal.
(ii) f(w1)=f(w2) for all w1,w2∈W, w1=w2. Otherwise, let u=f(w1)=f(w2) then by condition (1), u∈U. Let U′=(U∖{u})∪{w1,w2}, since f(U′)⊆V∪{u}, U′∩(V∪{u})=∅, we have U′∩f(U′)=∅. It is a contradiction to ∣U∣ being maximal.
Let W={w1,w2,…,wm}, ui=f(wi), 1≤i≤m then by (i) and (ii), u1,u2,…,um are distinct elements of U.
(iii) f(ui)=f(uj) for all 1≤i<j≤m. Otherwise, let v=f(ui)=f(uj)∈V, U′=(U∖{ui})∪{wi}, then f(U′)=V∪{ui}, U′∩f(U′)=∅. However, ∣f(U′)∣>∣f(U)∣. It is a contradiction to ∣f(U)∣ being maximal.
Therefore, f(u1),f(u2),…,f(um) are distinct elements of V. In particular, ∣V∣≥∣W∣. As ∣U∣≤39, we have ∣V∣+∣W∣≥61 and ∣V∣≥31. Let B=U∪W, then ∣B∣≤69 and f(B)∪B⊇V∪B=X. Overall, the desired smallest integer k is 69. □