Using multiple files or records about py-eddy-tracker

[AntSimi]

Yaml file for identification is an old one it will be removed at next realease. If you grid is a regular one you could use eddyid, but it can process only one file by command and only is first time step if you several in one.
To process a time series create a bash or somethings like that.
If your roms are unregular look at a previous closed issue.

[AntSimi]

Issue #5

[ivicajan]

Thanks a lot for swift reply!
Exactly what I needed for ROMS.
I was thinking about bash script and looping with date variable but then I might endup with bunch of daily files for Cyclonic and Anticyclonic outputs. If I want to make analysis for example on life duration of individual eddies I need to aggregate those daily outputs and have some dimension and then take/find ID of that particular eddy that might change from file to file, or there is already tool/script available?

Cheers
Ivica

[ivicajan]

Regarding ROMS inputs, I have already rotated and at collapsed to rho coordinates both u_eastward and v_northward components (Rutgers ROMS have the option). So, according to your #5, all I should do is just to make that speed norm aware of units (change units)? I could do it quickly using NCO ncrename for velocity unit attributes and put them to "meter second^-1"?

Cheers
Ivica

[AntSimi]

Thanks a lot for swift reply!
Exactly what I needed for ROMS.
I was thinking about bash script and looping with date variable but then I might endup with bunch of daily files for Cyclonic and Anticyclonic outputs. If I want to make analysis for example on life duration of individual eddies I need to aggregate those daily outputs and have some dimension and then take/find ID of that particular eddy that might change from file to file, or there is already tool/script available?

Cheers
Ivica

Atlas is produced in two step, identification first and tracking after. For tracking you must used EddyTracking with a yaml file, and also choose method to track.
To subset atlas you could use eddysubsetter. You could also take a look at the gallery from documentation.

[AntSimi]

Regarding ROMS inputs, I have already rotated and at collapsed to rho coordinates both u_eastward and v_northward components (Rutgers ROMS have the option). So, according to your #5, all I should do is just to make that speed norm aware of units (change units)? I could do it quickly using NCO ncrename for velocity unit attributes and put them to "meter second^-1"?

Cheers
Ivica

I think you need just to have correct units. Speed norm was needed in previous case, because u/v used another dimensions than h.