time series analysis - correlation

3 views (last 30 days)
Richard
Richard on 4 Apr 2012
My task involves finding the correlation between different variables for example air temperature and water temperature. The correlation can easily be calculated with the corrcoef command. However, when reading text books concerning time series analysis, they mention fitting models to the data to see if fluctuations in one variable i.e. temperature will affect the other (water temperature). I am rather overwhelmed by the amount of information provided in the literature, and dont really know where to start. Could anyone point me in the right direction of what I should be looking at doing first! Thanks in advance.

Answers (2)

tang
tang on 4 Apr 2012
Hi lestyn, Firstly,i'm so sorry that my english is poor that i am not sure to fully express my idea to you,anyway,i can try. Generally,if we want to know about the correlation between two seriess,we can draw the scatterplot chart first,then we should search for which model can fit the data best,well,some common models you should know,such as linear model,exp model,log-log model and so on.Accomplish these,some tatistics tests is a must,include t,F .etc MATLAB has a Statistics toolbox ,it can easily solve your problem. Some command you may use:scatter,regress . If you have any other problems,you can email it to me,my email is:beacher1988@163.com best regards.

Rajiv Singh
Rajiv Singh on 4 Apr 2012
If there is dynamic/causal dependence, you should try dynamic model fitting tools like those in System Identification Toolbox. See, ARX function for example.

Community Treasure Hunt

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

Start Hunting!

Translated by