On Tue, 10 Oct 2017, Sven Schreiber wrote:
Am 10.10.2017 um 14:08 schrieb Riccardo (Jack) Lucchetti:
> On Tue, 10 Oct 2017, Sven Schreiber wrote:
>> a student alerted me to the fact that it currently seems impossible to
>> save the output from the 'hurst' command into a variable for scripting
>> use.
>> The natural way would be a $-accessor I guess.
>
> In the short run, probably ($coeff being the obvious candidate, although
> the hurst command doesn't create a model struct). In a broader perspective,
> I guess we may want to provide a hurst() function.
Oh, and BTW, while you can get the _test_ results from the fractional
integration tests done with 'fractint', it doesn't seem possible to get the
_estimated_ order of integration from the local whittle or GPH approach,
right? (The ones which are also printed out after fractint.)
So instead of just hurst maybe we want to have a fracorder() function with
options "hurst", "whittle", "GPH", or something like that.
That's a nice idea, consolidating the similar pieces of
functionality. Shouldn't be difficult to arrange, I think.
Allin