3
$\begingroup$

I have created couple of graphs in R

d5 <- scan('d500.txt')
p1 <- acf(d5,lag.max = 36 ,type = c("covariance"),plot = TRUE)

enter image description here

Then the next one is correlation

p1 <- acf(d5,lag.max = 36 ,type = c("correlation"),plot = TRUE)

enter image description here

And pacf

p1 <- pacf(d5,lag.max = 36 ,plot = TRUE)

enter image description here

If I take a one difference operation

d51 <- diff(d5)
p1 <- acf(d51,lag.max = 36 ,type = c("correlation"),plot = TRUE)

enter image description here

And then again acf enter image description here

and pacf

enter image description here

How to interpret this?What does sinusoidal acf(cov) shows? Original data set is here

https://www.dropbox.com/s/8cmv02xtyvyrilp/d500.txt?dl=0

Improve this question
$\endgroup$
2
  • $\begingroup$ Acf's , pacf's etc reflect descriptive symptoms . Causes are found by analytics . Please post the original data so I can (maybe) unravel thus conundrum . $\endgroup$
    – IrishStat
    Commented Dec 27, 2016 at 16:35
  • $\begingroup$ also please advise what kind of data is this .. frequency of measurements $\endgroup$
    – IrishStat
    Commented Dec 27, 2016 at 18:04

1 Answer 1

Reset to default
4
$\begingroup$

You originally had 500,000 readings, 128 per second for 3,907 seconds. After taking the readings at 1 second interval (N=3907) the plot is here enter image description here . AUTOBOX identified a break point in parameters at period 2343 +/- 300 time points. Taking the most recent 1565 values this is then the identified model enter image description here (1,1,0) with an ACF of the residuals here suggesting model insufficiency enter image description here. It appears that there is significant auto-projective structure at period 60 suggesting an hourly effect .There were a few (8) anomalous data points ( also visually obvious) which were identified and rectified/adjusted . Some sort of SARIMA might be necessary of the form (1,1,0)(1,0,0)60 . I will sharpen my pencil and try to resolve this as ARIMA modelling is an ITERATIVE PROCESS.

AFTER ADDING A 60 PERIOD COEFFICIENT :

enter image description here

this captures the short-term (second) and the long-term (hourly) structure

Improve this answer
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.