2.2. Install pyCSAMT

This page takes you from an available Python interpreter to a verified core pyCSAMT installation. The core package is enough for the first survey workflow in this section. You can add optional features later, when a workflow actually needs them.

For supported versions, every optional dependency group, source and conda installs, application commands, compiled solvers, and troubleshooting, use the complete installation reference.

2.2.1. Check Python

pyCSAMT requires Python 3.9 or later. Python 3.11 or 3.12 is a practical choice for a new environment because both versions are covered by the test suite and are widely supported by scientific Python packages.

Check the interpreter that will create the environment:

python --version

If that command reports an older version, install a supported Python release before continuing.

2.2.2. Create An Isolated Environment

An isolated environment keeps pyCSAMT and its scientific dependencies from changing packages used by other projects. Create one inside your project or working directory.

On Linux or macOS:

python -m venv .venv
source .venv/bin/activate
python -m pip install --upgrade pip

On Windows PowerShell:

py -3.11 -m venv .venv
.\.venv\Scripts\Activate.ps1
python -m pip install --upgrade pip

Your shell prompt will usually show (.venv) after activation. Confirm that python and pip now belong to the same environment:

python -m pip --version

2.2.3. Install The Core Package

Install the stable package from PyPI:

python -m pip install pycsamt

Using python -m pip ties the installer to the active interpreter and avoids a common problem in which pip installs into a different Python environment.

2.2.4. Verify The Installation

First verify the Python package:

python -c "import pycsamt; print(pycsamt.__version__)"

The command should print a version without a traceback. Then verify the command-line interface:

pycsamt --help

The help output should list the available command groups. If pycsamt is not found, reactivate .venv and confirm the install location with:

python -m pip show pycsamt

Do not continue until both the import and command-line checks succeed. This prevents environment problems from being mistaken for data-loading or survey errors later in the workflow.

2.2.5. Add Optional Features Only When Needed

The core installation is the recommended starting point. Add an extra only when the next workflow requires it:

Need	Command	Extra
Coordinate transforms and map layers	python -m pip install "pycsamt[geo]"	geo
PyTorch-backed AI workflows	python -m pip install "pycsamt[torch]"	torch
Desktop and web applications	python -m pip install "pycsamt[app]"	app

These are examples, not the complete dependency matrix. See the optional feature groups in the installation reference before setting up TensorFlow, LLM providers, documentation tools, a full development environment, or an external inversion solver.

2.2.6. Continue With Your Data

Your environment is ready when:

• the virtual environment is active;

• importing pycsamt prints a version;

• pycsamt --help displays the command groups.

Continue to Identify your data format to identify the files from your instrument or processing system. After that, Configure a first session establishes output and style defaults, and Inspect a first survey loads and checks a first survey.

If you are contributing to pyCSAMT rather than installing it for use, follow the source-install instructions in Installation and then read Contributing.