Unevenly Sampled Time Series

jmr65

New Member
#1
Hello!

I am new to time series analysis and was hoping someone would be able to provide some suggestions. I study transportation data, and would like to perform a time series analysis on a data set made up of vehicle speed and the timestamp of when the data point was recorded. I'm using JMP software (which I am also relatively new to) to create my models.

The major problem I'm running into is that my data is unevenly sampled. It has been suggested before that I combine several data points into groups and use the median value, but I don't want to do this. I would like to analyze the data as is, so that I don't lose any data points...I just think that creating the groups would change the data too much.

So, the question is, can anyone suggest a method for creating the model that can handle the unevenly sampled data? I have looked at several resources, but they all seem to technical to be helpful to me (I'm an engineer with a very limited statistics background), so maybe there are some resources you could recommend that might help. I have also considered using the Transfer Function analysis in JMP, which seems like it might work. If anyone is familiar with this and could provide me with some basic guidelines of the process (looking for a better explanation than the help file), that would be much appreciated!

Thanks so much for your help! :)