For , all frequencies satisfy . At , the increment is . If , the first terms have and contribute at least , while the remaining tail has magnitude at most . This proves a lower modulus of continuity bound of order for small . The inverse theorem for trigonometric approximation supplies the matching upper bound because the positive lacunary trigonometric series has .
The original PDF prints , with a positive exponent. That premise is insufficient; the intended decay rate is . Indeed every bounded continuous periodic function has . For an explicit counterexample, on set
and extend periodically. This is continuous, but for , which is not for any . Thus the sign correction is required in the actual printed question, not just in its converted TeX.
For real continuous -periodic , write for best uniform approximation by degree-at-most- trigonometric polynomials and . A quantitative inverse theorem for trigonometric approximation is
where is universal. Here is a direct reason for the sum. Choose best approximants and . Write
The Bernstein inequality for trigonometric polynomials gives
For , monotonicity of best errors bounds by . The first terms are bounded by a fixed multiple of . Also . Now
proves the theorem, for example with a sufficiently large absolute constant .
Suppose, as intended, for , after enlarging to include any finite exceptional indices, where . Since , the inverse theorem gives
For take . Then , so monotonicity of the modulus of continuity and give . For , the separate bound covers the remaining scales. Therefore the corrected conclusion is
Now put . The given best approximant to the Weierstrass function leaves the uniformly convergent tail . Its supremum norm is exactly , attained at ; hence
The corrected inverse theorem gives . To show that this order is sharp, set . Every term of is nonnegative, and for the cosine equals because is odd. Thus
Given small , select with . Then . Combining the estimates gives the precise small-scale order
The exact approximation error can also be verified independently of the supplied fact: at the tail has alternating values . There are enough such points in one period for the trigonometric Chebyshev alternation theorem, since . This is the positive lacunary trigonometric series mechanism behind the best partial sums.
The coefficient sum is finite, so the Weierstrass M-test makes the cosine series uniformly convergent and continuous. Put and . The indicated partial sum belongs to because its largest frequency is . Its tail satisfies
and equality holds at . Thus its supremum norm error is exactly .
At the points , , every omitted frequency is an odd integer multiple of . Indeed for ,
Uniform convergence permits summing these values, giving
These are cyclically alternating extrema. Since , there are of them, so the trigonometric Chebyshev alternation theorem proves that the partial sum is the unique best approximant. Consequently the positive lacunary trigonometric series has
The positivity of every tail coefficient makes the same tail attain its full coefficient sum with alternating signs; a triangle bound alone would not establish optimality.
The Weierstrass M-test gives uniform convergence and continuity of the series. For the proposed trigonometric polynomial, set and . The tail satisfies
At , for , the integer is odd for every omitted frequency. Thus and . In particular .
Since , these points include at least consecutive alternating extrema in a single period. Also , so . The trigonometric Chebyshev alternation theorem proves that the proposed partial sum is the unique best approximant, with
This is an instance of positive lacunary trigonometric series: all omitted odd frequency ratios align their extrema.