0

I am trying to determine the correlation between an observed and theoretical set of data. Both sets of information are in the form of a table of deflection against time. Both samples have the same start and end time.

The problem i am encountering is with uneven samples sizes. for my theoretical result i have 300 results with a step size of 0.133ms and for my actual i have 400 results with a step size of 0.1ms. is there a simple way to get a coefficient bewteen these 2 sets without altering the sample size to a common value for both ?

Scott
  • 1
  • 1
  • 1
  • You mean you have X(t) for t=t0,to+0.133,t0+0.266,...,t0+299*0.133 and Y(t) for t=t0,t0+0.1,t0+0.2,...,t0+39.9 , and want to calculate the correlation between the two time series? – user31264 Oct 20 '14 at 18:51

1 Answers1

1

The answer to this question could range from a rather simple imputation of 'extra' values inserted at random to the smaller dataset to much more complex eg http://pascal.iseg.utl.pt/~jcaiado/Papers/Paper_LSSP0213_Revised.pdf Btw, I'd reclassify this question as 'unequal step size in time series' rather than unequal sample size. Also, this cross-correlation function discussion might be helpful Correlation between two time series

Community
  • 1
katya
  • 2,084
  • 8
  • 11