GRI Obs

Typed observation objects for geolocation measurements.

Overview

gri-obs provides self-contained observation objects that carry measurement data, noise characterization, and sensor geometry for geolocation systems. Each observation embeds everything needed to evaluate measurement functions against an emitter state vector.

Observation Types

Installation

pip install gri-obs

Or for local development:

git clone https://gitlab.com/geosol-foss/python/gri-obs.git
cd gri-obs
uv sync

uv sync creates the virtual environment, installs dependencies, and wires up the git hooks automatically (it sets core.hooksPath to .githooks). See CONTRIBUTING.md for the full workflow.

Usage

from gri_obs import TdoaObs, FdoaObs, AoaObs, RangeObs
from gri_pos import Pos
import numpy as np

# TDOA observation
tdoa = TdoaObs(
    time=100.0,
    value=1.5e-6,           # 1.5 microseconds
    std_dev=1e-9,            # 1 nanosecond
    collector1=Pos.LLA(38.0, -77.0, 400_000.0),
    collector2=Pos.LLA(39.0, -76.0, 400_000.0),
    sensor_id="tdoa_12",
)

# AOA observation from azimuth/elevation
aoa = AoaObs.from_azel(
    time=100.0,
    azimuth_rad=0.785,       # 45 degrees
    elevation_rad=0.524,     # 30 degrees
    std_dev_az_rad=0.001,
    std_dev_el_rad=0.001,
    collector=Pos.LLA(38.0, -77.0, 100.0),
    sensor_id="aoa_1",
)

# Filter mixed observation lists
from gri_obs import filter_tdoa, filter_aoa
all_obs = [tdoa, aoa]
tdoa_only = filter_tdoa(all_obs)

Timestamps

Observations accept any timestamp type -- float seconds, datetime objects, or gri_nsepoch.Time for nanosecond precision. The observation carries the timestamp but does not interpret it.

Dependencies

• gri-utils -- coordinate conversions
• gri-pos -- position objects
• gri-memoize -- AOA memoized properties

License

MIT -- see LICENSE.