0
$\begingroup$

I think this is a basic question but really I just need a definite starting point. I am working in R. I have a dataset as follows

rowid    TimePoint1   TimePoint2    TimePoint3    TimePoint4   TimePoint5   etc.
1          54            34           55           74             474    
2          36            63           83           344            636    
3          151           61           1643         34             976
etc.

(There are actually 10 TimePoints and hundreds of rows)

I want to basically create a best fit line (likely non-linear) which I can then compare to another group with the same dataset structure.

I think I really need to know what this type of analysis might be called- firstly to create the best fit (is it just regression?) and then to compare with another group.

Improve this question
$\endgroup$
2
  • $\begingroup$ If the aim is to compare a group of observations to another group, maybe a full graphical comparison of the two sets, showing each data line, would be the most effective. Another thing I would think of is a simple nonparametric smoother. $\endgroup$
    – lambda_vu
    Commented Nov 29, 2016 at 9:10
  • $\begingroup$ Searching for "repeated measures" may help. If the data you gave are real, then you have lots of problems: difficult to know how the third case could be modelled! $\endgroup$
    – Nick Cox
    Commented Nov 29, 2016 at 9:43

0

Reset to default

Browse other questions tagged or ask your own question.