Let Z be a set of bit strings in which number of 1-bits is more than number of 0-bits. Consider a function on bit strings f(X) = (the number of 1-bits in X) – (the number of 0-bits in X), where X be an arbitrary bit string: X=(x1, x2, …, xN). So Z is a set of bit strings X that f(X) >=1. We prove that A = Z.
Statement (1). If XA then XZ.
Prove this statement by induction over N - the length of the string X.
1.1. Case N=1. As f(0) = -1, f(1) = 1, therefore, 1 Z, 0 Z. Clearly, 0 A, and, by condition, 1A, so the base of induction is proved.
1.2. Let the statement (1) is proved for any string X A of the length less than N. Consider a string X of the length N. By condition, X is of one of the form: X = aB or X = 0aB or X = a0B or X = aB0, where aA, BA. Clearly, that the length of both strings a and B is less than N. By induction, f(a)>=1 and f(B)>=1.
If X=aB then f(X) = f(a)+f(B)>=2, and, hence, XZ.
If X = 0aB or X = a0B or X = aB0 then f(X) = f(a)+f(B) -1 >=1, and, hence, XZ.
So, the statement (1) is proved for strings of length N, and, by induction, for strings of any length.
Statement (2). If XZ and f(X) >= 2 then there exist strings aZ and BZ that X=aB.
2.1. Let n be a minimal index that f(x1, x2, …, xn) >= 1. If xn = 0 then f(x1, x2, …, xn-1)>=2 and this contradicts to condition that n is a minimal index that f(x1, x2, …, xn) >= 1. Therefore, xn = 1.
If f(x1, x2, …, xn) >= 2, then f(x1, x2, …, xn-1) = f(x1, x2, …, xn) -1 >=1 and this contradicts to condition of the minimality n. Therefore, f(x1, x2, …, xn) = 1.
Let a = (x1, x2, …, xn), B = (xn+1, xn+2, …, xN). Then X = aB, f(a) =1 and f(B) = f(X) – f(a) >= 1.
So, aZ, BZ and X=aB and the statement is proved.
Statement (3). Any string X Z is an element of A.
Again, the proof will be based on induction on N - the length of the string X.
3.1. If N=1, f(X)>=1 then X=(1), and XA by condition.
3.2. Let the statement (3) is proved for any string XA with the length less than N. Consider several cases:
(a) If f(X)>=2 then, by statement (2), there exist strings aZ and BZ, that X=aB. As the length of a and B less than N, by induction, they are elements of A. Then X = aB belongs to A by condition.
(b) If f(X)=1 and x1=0 then f(x2, …, xN)=2 and, by statement (2), there exist strings aZ and BZ, that (x2, …, xN)=aB. As the length of a and B less than N, by induction, they are elements of A. Then X = 0aB belongs to A by condition.
(c) If f(X)=1 and xN=0 then f(x1, …, xN-1)=2 and, by statement (2), there exist stringsaZ and BZ, that (x2, …, xN)=aB. As the length of a and B less than N, by induction, they are elements of A. Then X = aB0 belongs to A by condition.
(d) If f(X)=1 and x1=xN=1 then f(x2, …, xN-1)=-1.
Let n>=2 be a minimal index that f(x2, …, xn) <= -1. Then f(x2, …, xn-1) >=0, xn=0 and f(x2, …, xn-1) =0.
Let a =(x1,…,xn-1) and B = (xn+1,…,xN). Then X = a0B, f(a)=1 and f(B) = f(X)-f(a)+1 = 1.
By induction, a and B are elements of A. Then X = a0B belongs to A by condition.
As (a)-(d) form a complete system of cases, the statement (3) is proved for strings of the length N, and, by induction, for any strings.
It follows from (1) and (3) that A=Z, that is, A is a set of bit strings in which number of 1-bits is more than number of 0-bits.
Answer. A is a set of bit strings in which the number of 1-bits is more than the number of 0-bits.
Comments