The updated data or outdated sample website?

[Zonca-Bushnell]

I used the sample program, which requires data in ZARR format, but I only found NC format data in the relevant dataset. This makes it difficult for me to understand how to use the function. Could this be because the example documentation I referenced is outdated?
(https://pyeddytracker.readthedocs.io/en/stable/python_module/02_eddy_identification/pet_filter_and_detection.html#sphx-glr-python-module-02-eddy-identification-pet-filter-and-detection-py)

Additionally, I'd like to ask:
Can I use nc files to track the trajectory of a single vortex over an extended period (requiring analysis of its lifecycle)?

[4Kalil]

I am a noob, and I'm having the same problem as follows:

1. I see the
   py_eddy_tracker_sample.get_demo_path("eddies_med_adt_allsat_dt2018/Anticyclonic.zarr")
   in the code (in "Track animation" part), but the "get_demo_path" module is unable to use, and I also cannot find any .zarr file in the program.

Was it because of the py-eddy-tracker version that I download? I get it from the git link.

2. Now I can detect eddies in a single day, and then save it to a .nc file, but I cannot figure out how to detect eddies for a series of time such as one month without a sample .zarr file. Therefore, I cannot move on through this program, such as tracking the eddies and forming EddyID, etc.

I wonder that is it necessary to form a .zarr file to save infomations about eddies for a period of time? In my point of view, esp. as a noob, I think it is.

To sum up, I'm wondering that would you please send me some sample files in .zarr format, such as "eddies_med_adt_allsat_dt2018/Anticyclonic.zarr" in your code? It would be very thankful for you to do that.

If it wasn't convinient to publish it on github, would you please send it to my email address?
That is: kal2156589077@gmail.com
Thank you very much for your program!

[AntSimi]

@Zonca-Bushnell zarr and netcdf are 2 available format and you could use both, there is missing sign in your link, right one is here

[AntSimi]

@4Kalil get_demo_path is only a method to get demonstration dataset to produce example. To load your own one use following method TrackEddiesObservations.load_file.
To run tracking you need to do identification on each time step of your interest and after run tracking on your identification file: default tracking

[Zonca-Bushnell]

I have attempted to use PET to track the trajectories of individual eddies within a multi-day nc-Eddy dataset.
In my expectation, each eddy should possess a unique ID and correspond to a distinct vortex. However, the actual outcome is that the number of tracks matches the number of days. That is, it assigns a single ID to encompass all vortices occurring on a given day.
Consequently, my query is: what is the ID generation logic employed by PET's tracking functionality?

Naturally, I also suspect that my NC files may contain issues with the time variable. As I was unable to locate data in the example set that tracked multiple vortices over several days, I have instead utilised NC files that I have merged myself.

Therefore, my second question is: does my ID generation issue stem from PET's strict constraints on the TIME field within multi-day data?