Offline Installation

SudaPy supports installation in environments without internet access. This is useful for field work or restricted networks common in surveying and government contexts.

Strategy overview

Method	Best for	Complexity
Wheelhouse (pip)	Pure Python + pre-built wheels	Low
conda pack	Full environment with C libraries	Medium

Method 1: Wheelhouse (pip)

On a machine with internet

# Clone and build the wheelhouse
git clone https://github.com/Osman-Geomatics93/sudapy.git
cd sudapy
python scripts/build_wheelhouse.py

This downloads all wheels into a wheelhouse/ directory. Copy the entire wheelhouse/ folder to a USB drive.

On the offline machine

pip install --no-index --find-links=wheelhouse/ sudapy

Warning

Pip wheelhouses work best when the online and offline machines have the same OS and Python version. For geospatial C libraries (GDAL, PROJ), conda is more reliable.

Method 2: conda environment

On a machine with internet

# Create and export the environment
conda env create -f environment.yml
conda activate sudapy

# Pack the environment for transfer
conda install -c conda-forge conda-pack
conda pack -n sudapy -o sudapy-env.tar.gz

On the offline machine

# Unpack
mkdir -p ~/envs/sudapy
tar -xzf sudapy-env.tar.gz -C ~/envs/sudapy

# Activate
source ~/envs/sudapy/bin/activate   # Linux/macOS
# or
~/envs/sudapy/Scripts/activate.bat  # Windows

# Fix prefixes
conda-unpack

Verify offline installation

sudapy doctor
sudapy info
sudapy crs list

All three commands should work without network access since they use only local data.