UnaryMeasurement - GTSAM Docs

Created by Codex.

Store a typed measurement on one key, together with its uncertainty.

Authors: Frank Dellaert, et al. (see THANKS for the full author list)

See LICENSE for the license information

try:
    import google.colab
    %pip install --quiet gtsam-develop
except ImportError:
    pass

import gtsam
import numpy as np

from gtsam import symbol_shorthand

C = symbol_shorthand.C
K = symbol_shorthand.K
P = symbol_shorthand.P
S = symbol_shorthand.S
X = symbol_shorthand.X

Mathematical idea¶

A unary measurement record packages a key, typed observation, and covariance,

(i,z_i,\Sigma_i).

(1)

It is intentionally model-free: a consumer may later define a residual such as $h(x_i)-z_i$ , but the record itself stores no prediction function and contributes no optimization error.

When to use it¶

UnaryMeasurement<T> is a lightweight measurement record rather than an error-bearing nonlinear factor. TrajectoryAlignerSim3 uses it to associate timestamp or frame keys with measured poses.

Python exposes UnaryMeasurementPose3, UnaryMeasurementRot3, and UnaryMeasurementPoint3.

pose_noise = gtsam.noiseModel.Isotropic.Sigma(6, 0.05)
measured_pose = gtsam.Pose3(gtsam.Rot3(), np.array([1.0, 2.0, 3.0]))
measurement = gtsam.UnaryMeasurementPose3(X(4), measured_pose, pose_noise)

print("key:", measurement.key())
print("translation:", measurement.measured().translation())
print("noise sigmas:", measurement.noiseModel().sigmas())

key: 8646911284551352324
translation: [1. 2. 3.]
noise sigmas: [0.05 0.05 0.05 0.05 0.05 0.05]

Practical notes¶

The key is what joins measurements across trajectories. TrajectoryAlignerSim3 only uses child measurements whose keys also appear in the parent trajectory, so consistent key assignment is essential.

Source¶

UnaryMeasurement.h