Annual trend of uneven timeseries

I would like to know how to retrieve the annual trend of an uneven timeseries, does anybody have an answer?

My timeseries is made up by two vector: time vector and the related data vector.

This is an example of the time vector: T=['03-Nov-1994' '28-Dec-1994' '06-Jan-1995' '26-Feb-1995' '07-Jan-1996' '13-Feb-1996' '26-Jan-1997' '14-Feb-1997' '23-Nov-1997' '09-Jan-1998' '28-Feb-1998' '05-Jan-1999' '07-Jan-2000' '06-Jan-2001' '21-Feb-2001' '06-Jan-2003' '24-Dec-2003' '01-Jan-2005' '03-Jan-2006' '05-Feb-2007' '16-Jan-2008' '28-Jan-2010' '13-Jan-2012' '14-Feb-2014' '16-Jan-2016' '31-Dec-2016' '07-Jan-2020' '25-Dec-2020' '08-Jan-2022']

I don't want to temporally interpolate my data, because in that way I would lose my data variability.

I think a good solution could be a spectrum analysis, but I am not sure on how to do it.

[dpb]

"retrieve the annual trend of an uneven timeseries,"

Well, it would help if you could define what you think the "annual trend" would be.

And, of course, for folks to try things it would be much more useful to have a representative dataset that illustrates the nature of the data as well.

[Antonino Ferola]

Sorry, i want to cumpute the trend over 30 years of my time series and therefore express it in terms of variation of my variable per year.

I attached the Time Vector and a Data Vector.

Thanks in advance!

[dpb]

Same words as before -- what is "trend"? A linear trend computed annually, maybe? That'd be simple enough with the year as the grouping variable with the <splitapply> workflow with the functional being as simple as polyfit or detailed to use <fit> of the Curve Fitting TB if want all the statistics to go with it.

"In terms of variation of my variable per year" is also a pretty nebulous description; you could standardize the slope by the mean, pehaps,is what you're asking for?

So, what's this? Nothing in a time vector that can be recognized as years....certainly nothing to be able to do anything with by a year...