mean trend of 5 time series regressions

1 vue (au cours des 30 derniers jours)
aine gormley
aine gormley le 31 Oct 2018
Commenté : aine gormley le 1 Nov 2018
Hi, so I have completed 5 x linear regressions (same variable on same time series, just measured in 5 different ways). Now I am trying to compute the mean trend (of the 5). Is it using the mean_slope = mean (slope etc.)? or can anyone point me in the right direction as I am going round in circles. Thank you for your time!
  2 commentaires
dpb
dpb le 1 Nov 2018
Why not just combine all the data and fit it?
If the sample space and weights are similar for the five measurements and the assumptions of similar noise distribution hold, then the two results will be essentially the same; there being a sizable difference would be indicative of something not being consistent amongst the datasets--of course, that probably would show up by a significant difference in estimates from one set to another as well.
aine gormley
aine gormley le 1 Nov 2018
Ah -sometimes the simplest solution is the right way! thanks a million for your time :-)

Connectez-vous pour commenter.

Réponses (0)

Catégories

En savoir plus sur Descriptive Statistics dans Help Center et File Exchange

Produits


Version

R2018b

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by