User Instructions

Steps to run ARMP

1. Create your project in ARMP/params/config.in

Set ARMP workflow flags and variables

2. Prepare input data

Prepare input file-pointer(as .in files) required to run ARMP in ARMP/data/ or user-defined dir_in

A list of AR object/tag NetCDF files are specified in "{}_{}_tag_list.in".format(model, ARDT)

Optionally, for AR-related climate variable analysis e.g., precipitation, A list of climate data files need to be included in "{}_{}_clim_list.in".format(model, clim_var)

.in pointer files and .nc data files are not necessary in the same directory

3.Installation

install the package each time we create a new project, i.e., when config.in is modified.

in package root directory

pip install .

4. Run ARMP

ARMP is driven by the driver*.py scripts

$ cd ARMP
$ python driver_AR.py

If more climate data analysis is desired and the processed AR tag files are already in output or dir_out,

set restart = True and include_clim = True

and

run driver_clim.py directly

$ python driver_clim.py

5. AR characteristics with BlobStats

Due to a Known issue for installing BlobStats as a python env dependency, the current version of the package require users to install and execute BlobStats offline. see instructions in stats/blobstats.md

6. Metrics visualization

Follow the instructions printed on screen from ARMP run, go to graphics/ and custom and run desired plotting scripts.

code structure of ARMP:

stats/: statistics for AR objects and climate data

metrics/: metrics functions

app/: scripts for metrics workflow, i.e., calculating metrics/ of stats/

graphics/: plotting scripts, run as offline after executing ARMP