Upper limits

[stefanocovino]

Hi,

I know this is a question that should be better defined, at least referring to a specific statistical test. However, I wonder if you have thought to a way to include upper limits in the analysed time series.

Thanks,
Stefano

[leliel12]

Can you explain a little more your idea? i don't understand how/where/why you want an upper limit.

[stefanocovino]

I was cryptic. Sorry for that. What I mean is simply that it happens, in a time-series, to have upper limits (or even lower limits, in principle, if we are not speaking of typical astronomical time series). Depending on the specific task, upper limits can be meaningful for assessing a variability. However, I do not know how to indicate them in the context of your package, if there is a way. In any case it could be interesting to code a standard way to report these measurements. Stefano

[leliel12]

you mean to limit the "magnitude" of the time series?

[stefanocovino]

Let's assume, for instance, we have ten magnitude measurements. And nine of them are just upper limits, i.e. nothing is detected. One of them gives a measurement brighter than the limits. This immediately says there is important variability, and of course, with ten measurements everything is trivial. The idea would be to have tools to assess variability able to compare actual measurements with limits. Stefano Il giorno giovedì 14 marzo 2019, Juan BC <[email protected]> ha scritto:

…

you mean to limit the "magnitude" of the time series? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#10 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AA1cHAgx-tUK3apdS7Wpn2syzzA3ugBkks5vWkH8gaJpZM4btmdb> .