I have data that is vaguely sinusoidal/periodic - there is variation in different periods, but in more than amplitude and the length of a period. I've been told to use pca to figure out the primary contributing variables to the variation, but I'm not sure how to do that with this data? I mean, it's in two dimensions right now.

  • 1
    $\begingroup$ Welcome to our site, Megan. What are the "two dimensions"? Time and a response? Two paired responses? It is unclear what you mean by "variation": could you be more explicit about what this amounts to? $\endgroup$ – whuber May 28 '15 at 22:02
  • $\begingroup$ Yeah, it's time and pressure. As to variation - if I divide it into periods and plot them against each other, it's clear that parts of the curve are sometimes significantly differently curved (ie: more table-shaped, or more concave towards the end of the period) $\endgroup$ – Megan May 29 '15 at 2:11
  • $\begingroup$ It's going to be difficult for anyone here to speculate on what your advisor might have been thinking. Why not ask them to clarify? $\endgroup$ – whuber May 29 '15 at 13:17

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.