The first omitted term controls the error
Suppose positive magnitudes decrease to zero and the signs alternate. After retaining a finite initial block of terms, the absolute remainder is at most the magnitude of the next term. Consecutive partial sums bracket the limit, which explains both the size bound and the remainder's sign.
Count terms before solving an inequality
For , the next term is . If the series starts at and terms are retained, its last included index is and the first omitted index is . An index and a term count need not be the same number.
New example: a geometric alternating sum
Consider . The magnitudes have ratio . After terms, the alternating-series estimate gives
For an error at most , require . Since and , five terms are sufficient by this bound.
Here the geometric formula supplies an independent exact check: and . Four terms have error , while five have error .
A guarantee is not an exact error
If the bound is below the tolerance, the approximation is certified. If the bound is above it, the test alone is inconclusive about that approximation. Proving the smallest possible term count may require a lower bound or an exact remainder, as in the example.
Keep the inequality in the question: “at most” allows equality, while “less than” is strict. Finally, display rounding is a separate error source; retain enough digits if a decimal approximation must itself satisfy the requested tolerance.
Related question
Apply this knowledge
Use the concept guide to understand the reasoning, then return to the complete question and worked answer.
Alternating Series Error Bound: How Many Terms Approximate cos 1?Sources
These references support the core concepts and interpretation boundaries explained above.