Let
Is differentiable at ?
Locate x = −1 relative to the interval of convergence. The ratio test gives radius , and at both endpoints converges absolutely (it is dominated by ). So is defined on the closed interval , and is an endpoint — which is exactly why the question is not automatic. Inside differentiability is free; at an endpoint it must be argued.
Differentiate term by term inside the interval. For ,
using with and then dividing by .
Test the differentiated series at the endpoint. Putting into the derivative series gives
The alternating harmonic series converges (by the alternating series test) to . This is the decisive difference from the case , where the same series becomes the divergent harmonic series and the derivative fails to exist.
Upgrade convergence to differentiability with Abel's theorem. Since the differentiated series converges at , it converges uniformly on , and a uniformly convergent series of derivatives may be integrated/differentiated term by term up to and including the endpoint. Hence is differentiable at (one-sidedly from within , and two-sidedly if one uses the analytic continuation of past ) with
Confirm with the closed form. The formula is continuous at and gives
agreeing with the series value. (For reference, .)
Contrast the two endpoints. At the difference quotient tends to , so has a vertical tangent there and is not differentiable. The single extra factor of at turns a divergent series into a convergent one — the whole answer turns on that.
Need to solve a different problem like this? Open the solver →