Convert GeoTiff to DSS Grid Record

This repository contains a utility to read a GeoTiff (single band) and write the data to a DSS grid record.

Local Development

Build and Run

The shell script docker_run.sh can be used to build/run the Docker image/container. There are argument options for running tests, keeping the container alive, or building binaries. Any combination of arguments can be used but the one argument (dev) is used mainly for local development. Below are the arguments and what they do.

Argument	Evaluated In	Description
dev	docker_run.sh	Run the container interactively in bash with a mounted volume
itest	entrypoint.sh	Integration testing using hec-dssvue.sh; verbosity defined in each script
utest	entrypoint.sh	Unit testing the tiffdss shared object
build	entrypoint.sh	Run 'make' for tiffdss
alive	entrypoint.sh	Keep the container alive
help	entrypoint.sh	Print help message

Build tiffdss

> ./docker_run.sh build

Usage: tiffdss binary

tiffdss -p [-c] [-d] [-g] [-h] [-m] [-n] [-s] [-u] <input_Tiff> <output_DSS>

Options

-c [UNDEFINED_COMPRESSION_METHOD | ZLIB_COMPRESSION]
    DSS compression method;  or  (default: ZLIB_COMPRESSION)

-d [PER-AVER | PER-CUM | INST-VAL | INST-CUM]
    DSS data type; (default: PER-AVER)

-g [HRAP | ALBERS | SHG | SPECIFIED_GRID_TYPE | UTM]
    DSS grid type; (default: ALBERS)

-h [N | S]
    Hemisphere N or S; use with -g UTM (default: N)

-n
    Time zone name (default: GMT)

-l
    zsetMessageLevel (default: 0 (None))

-m
    No data value (default: 9999)

-p "/a/b/c/d/e/f/"
    DSS pathname

-s [0 | 1]
    DSS grid record time stamped; (default: 1)

-u
    DSS grid record units (default: MM)

-z
    UTM Zone 1-60; use with -g UTM

Python Unit Testing

Integration Test

1. Converting GeoTiff(s) in tests/integration/fixtures to DSS7
2. Converting the DSS7 file to DSS6
3. GridInfo differences between DSS7/DSS6

fixtures JSON setup:

This example will test each GeoTiff in the fixtures/mrms directory. Each of the attributes must correspond to the GeoTiff metadata.

test_products.json

{
    "mrms": {
        "apart": "",
        "bpart": "mrms",
        "cpart": "precip",
        "fpart": "qpe",
        "data_type": "per-cum",
        "data_unit": "mm",
        "data_interval": 3600
    },
    "hrrr": {
        "apart": "",
        "bpart": "hrrr",
        "cpart": "precip",
        "fpart": "qpe",
        "data_type": "inst-val",
        "data_unit": "mm",
        "data_interval": 3600
    }
}

Unit Test

1. libtiffdss.so C function testing
2. Numpy arrays (zeros, ones, random, and missing) generated
3. Numpy array statistics (min, max, and mean) compared to shared object statistics

Unit testing against the libtiffdss.so shared object uses Python ctypes module. The shared object is moved to /usr/lib during the build process to simplify loading the library.