STAREMaster_py is the python implementation of STAREMaster. It is used to create sidecar files for a collection of remote sensing products

Install

With pip

The installer will install the create_sidecar_files.py script and its dependencies.

Create a virtualenv:

mkvirtualevironment staremaster_py

Install from pypi

pip install staremaster

we then can install STAREMaster_py with

pip3 install -e STAREMaster_py/

pyhdf issues

Installing pyhdf from pip does not always seem to be trivially possible. Simple workaround is using conda;

conda create --name staremaster
conda activate staremaster
conda install -c conda-forge pyhdf

Then:

pip3 install -e STAREMaster_py/

Requirements

• pystare
• pyhdf
• numpy
• netCDF4
• argparse
• xarray
• dask['distributed']
• filelock

xarray and dask are hardcoded dependecies, but will become optional in the future

Usage

usage: create_sidecar_files.py [-h] [--folder folder] [--files files [files ...]] [--grid files] [--out_path OUT_PATH]
                               [--product product] [--cover_res cover_res] [--workers n_workers] [--archive archive]
                               [--parallel_files]

Creates Sidecar Files

optional arguments:
  -h, --help            show this help message and exit
  --folder folder       the folder to create sidecars for
  --files files [files ...]
                        the files to create a sidecar for
  --grid files          the grid to create a sidecar for (e.g. IMERG)
  --out_path OUT_PATH   the folder to create sidecars in; default: next to granule
  --product product     product (e.g. cldmsk_l2_viirs, hdfeos, l2_viirs, mod05, mod09, vj102dnb, vj103dnb, vnp02dnb, vnp03dnb,
                        ssmi)
  --cover_res cover_res
                        max STARE resolution of the cover. Default: min resolution of iFOVs
  --workers n_workers   use n_workers (local) dask workers
  --archive archive     Create sidecars only for granules not listed in the archive file. Record all create sidecars and their
                        corresponding granules in it.
  --parallel_files      Process files in parallel rather than looking up SIDs in parallel

e.g.

python3 create_sidecar_files.py --workers 4 
       --product MOD09 --file ~/MOD09.A2019317.0815.006.2019319020759.hdf

Extension

To add support for additional products, we need the following:

1. a module in products/ containing
   • a class for the product that implements the reading of the geolocation and the gring
   • a method that implements the write_sidecar() function
2. an import of the new module in products/__init__.py
3. argument parsing and switch for the added product in create_sidecar_files.py