Closed michael135 closed 5 years ago
seasonal_strength1 and seasonal_strength2 are for multi-seasonal time series, representing seasonal_strength for the 1st and 2nd seasonal component.
If I choose just seasonal_strength, it means, that there will be just 1 kind of seasonality?
Yes.
Sent from my iPhone
On Mar 6, 2019, at 19:03, Michael Dym notifications@github.com<mailto:notifications@github.com> wrote:
If I choose just seasonal_strength, it means, that there will be just 1 kind of seasonality?
— You are receiving this because you commented. Reply to this email directly, view it on GitHubhttps://github.com/ykang/tsgeneration/issues/6#issuecomment-470065725, or mute the threadhttps://github.com/notifications/unsubscribe-auth/ABJU1i9SeEkzFhXnJ_4MLTyVrZ9Z5vucks5vT6BkgaJpZM4bgTSc.
Does it makes sence if I choose:
When:
Isn't it some kind of contradiction?
I don't think that is a realistic setting reflecting the real world.
Just to assure that we are on the same page, in order to create "realistic" seasonal time series.
There should be:
Or some additional features also should be greater than 0 ?
I think you just need to set seasonal_strength > 0
and leave seas_acf1
and seas_pacf1
as 0
(in the Shiny app, 0
really means unrestricted), unless you have some specific interest (target) of acf
and pacf
.
thank you @feng-li, it's very helpful information.
Without GUI, the logic is the same?
If I don't specify "some feature" it's not necessary 0
?
What is the difference between following features?
And in particular if seasonal_strength = 1 is it possible, that seas_acf1=0?
I looked through
tsgeneration paper
, but I didn't find. Please let me know if I miss something.