URFT - Universal Reference Frame Transformer

Overview

The Universal Reference Frame Transformer transforms coordinates between supported reference frames, including epoch handling for dynamic frames and optional uncertainty propagation.

This repository also includes:

• A batch file module (file_transformer.py) used for processing csv data.
• A Streamlit GUI (app.py) for interactive single-point and batch transformations.

Installation

To install URFT use pip

pip install urft

If you want to use the gui use following pip command

pip install urft[gui]

Supported Reference Frames

Transformations can be completed between any of the following frames.

Static

GDA2020, GDA94, AGD66, AGD84, MGA94, MGA2020

Dynamic

Australian:

ATRF2014

ITRF:

ITRF88, ITRF89, ITRF90, ITRF91, ITRF92, ITRF93, ITRF94, ITRF96, ITRF97, ITRF2000, ITRF2005, ITRF2008, ITRF2014, ITRF2020

WGS84:

WGS84 (Transit), WGS84 (G730), WGS84 (G873), WGS84 (G1150), WGS84 (G1674), WGS84 (G1762), WGS84 (G2139), WGS84 (G2296), WGS84 Ensemble

Dependencies

• GeodePy
• NumPy
• Shapely
• PyProj

For GUI:

• Streamlit
• pydeck

Core Functionality

Transformation can take place in any of the three coordinate types. Depending on which coordinate type is input will decide which function should be used.

XYZ (Cartesian)

universal_transform(
    x, y, z,
    from_ref, to_ref,
    from_epoch=None, to_epoch=None,
    plate_motion="auto", vcv=None,
    return_type="xyz", output_zone=None, 
    ignore_errors=False, verbose=False
)

LLH (Latitude/Longitude/Height)

universal_transform_llh(
    lat, lon, el_height,
    from_ref, to_ref,
    from_epoch=None, to_epoch=None,
    plate_motion="auto", vcv=None,
    return_type="llh", output_zone=None,
    ignore_errors=False, verbose=False
)

ENU (Projected Easting/Northing/Height + Zone)

universal_transform_enu(
    east, north, height, zone,
    from_ref, to_ref,
    from_epoch=None, to_epoch=None,
    plate_motion="auto", vcv=None,
    return_type="enu", output_zone=None,
    ignore_errors=False, verbose=False
)

Key parameters

• plate_motion: "auto" selects plate motion automatically; "aus" forces the Australian plate motion model.
• vcv: optional 3×3 NumPy array for uncertainty propagation.
• return_type: one of "xyz", "llh", "enu".
• output_zone: Will set output_zone if using "enu" return type.
• ignore_errors: When set to true will run transformation even if it shouldn't
• verbose: When set to true transformation path will be printed

Epoch Requirements

Epoch inputs depend on whether the source/target frames are static or dynamic.

• Static → Static: no epochs required
• Static → Dynamic: to_epoch required
• Dynamic → Static: from_epoch required
• Dynamic → Dynamic: from_epoch and to_epoch required

Output

Output can be in any of the three coordinate types. All functions return a dictionary containing:

• type: output coordinate type (xyz, llh, or enu)
• vcv: 3×3 NumPy array if uncertainties are available, else None
• coords: coordinate fields matching the output type

For example to access the coords from an llh coordinate type you would use to following

latitude = result['coords']['lat']
longitude = results['coords']['lon']

Streamlit GUI

A Streamlit app is provided in app.py with two pages:

• Single Point
• Batch Processing

To run:

streamlit run app.py

Batch / File Transformation Helper

file_transformer.py let you complete batch processing of points in a csv. To do so it has two classes to handle required data:

• TransformParams holds transformation parameters that are not the coordinates
• CSVCoordinateMapping for mapping coordinate columns in csv (supports xyz, llh, and enu)

Using these classes a csv file can be transformed using the following function.

csv_transformation(file_path, CSVCoordinateMapping, TransformParams)