Getting Started

This section is the shortest path from a clean Python environment to a first pyCSAMT survey workflow. It is written for users who want to install the package, understand which data format they have, configure a working session, and inspect a first field survey before moving into processing, inversion, or AI-assisted workflows.

Start here when you are new to pyCSAMT v2, when you are setting up a new machine, or when you need a reproducible first-pass workflow for a survey directory.

Fast Path

For a standard Python environment and a folder of EDI files, the complete first path is:

python -m pip install pycsamt
pycsamt --help
pycsamt edi info data/edis
pycsamt edi validate data/edis

Then in Python:

from pycsamt.api import read_edis
from pycsamt.pipeline import Pipeline

survey = read_edis("data/edis", strict=False)
print(survey)
print(survey.summary().df.head())

pipe = Pipeline.from_preset("basic_qc")
result = pipe.run(survey.to_collection(), outdir="results/first_survey")
print(result.summary())

If that works, pyCSAMT is installed, the EDI reader can see your data, the public API is available, and the processing pipeline can produce a first quality-control run.

For a condensed end-to-end demonstration — processing, Occam2D inversion, geological interpretation, and AI-based inversion in a few code blocks — see Quick Start.

Recommended Reading Order

Read the pages in this order for the smoothest first experience:

Step	Page	Use it to
1	Installation	Choose an environment, install core pyCSAMT or optional extras, and verify the CLI, API, AI backends, GUI, and documentation tools.
2	Data Formats	Identify whether your inputs are EDI, Zonge AVG/AMTAVG, J-format, spectra, TDEM/TEM, site tables, or inversion/model files.
3	Configuration	Configure API table behavior, plotting style, pipeline output, AI backend selection, and agent/LLM settings.
4	First Survey	Load a first EDI survey, inspect station metadata, check parser errors, build QC tables, save figures, and run basic_qc.

The order matters because pyCSAMT workflows become much easier when you know which optional dependencies are installed, which field format you are using, and where outputs will be written before you begin processing data.

What Each Page Covers

Installation

Installation recipes for PyPI, source checkouts, development, docs, geospatial workflows, desktop/web apps, AI backends, and LLM provider SDKs. It also includes verification commands and troubleshooting notes.

Data Formats

A practical guide to supported input families. It explains the standard EDI path, legacy/field formats, conversion routes into EDI, site metadata, inversion files, duplicate-station policies, and unit conventions.

Configuration

The core runtime settings users usually touch first: APIFrame versus pandas output, pipeline output directories, plotting defaults, AI backend selection, and agent configuration.

First Survey

The first end-to-end survey walkthrough. It demonstrates pycsamt.api.read_edis(), station summaries, parser-error review, quality tables, confidence plots, the basic_qc pipeline preset, and CLI equivalents.

Quick Start

A condensed tour of the core v2 workflow: EDI processing, 2-D inversion with Occam2D, geological interpretation and export, and AI-based 1-D inversion — each in a short, copy-paste-ready code block.

Choose Your Entry Point

Different users arrive with different needs. Use this table to jump directly to the right page.

If you want to…	Start with…
install pyCSAMT on a new machine	Installation
decide between venv, conda, core install, and full extras	Installation
enable the desktop GUI or Dash-based app components	Installation
enable PyTorch, TensorFlow, or LLM-backed agents	Installation and ../agents/llm_configuration
understand whether your field files are supported	Data Formats
convert AVG, J-format, spectra, or TEM/TDEM data into an EDI workflow	Data Formats
configure where pipeline results and figures are written	Configuration
load an EDI directory and inspect stations	First Survey
run the first quality-control pipeline	First Survey

Before You Continue

Before moving into tutorials, pipeline customization, or inversion preparation, confirm these basics:

• python -c "import pycsamt; print(pycsamt.__version__)" prints a version.

• pycsamt --help shows the command groups.

• You know whether your input data are EDI, AVG, J-format, spectra, TDEM/TEM, site metadata, or model/inversion files.

• Your output directory is explicit, for example results/first_survey.

• read_edis(..., strict=False) loads the expected number of EDI stations.

• Parser errors are understood before processing results are trusted.

• The first basic_qc pipeline run writes a report and plots.

Where To Go Next

Once the getting-started workflow is working, continue with:

• Read an EDI Survey for a deeper EDI loading tutorial;

• Inspect and QC a Survey for richer survey diagnostics;

• ../pipeline/index for reproducible processing pipelines;

• ../pipeline/presets for built-in workflow presets;

• Prepare an Occam2D Inversion for a first inversion preparation path;

• ../agents/overview for AI-assisted survey review, processing plans, and result summaries;

• Development if you are contributing to pyCSAMT itself.

Getting Started Pages

• Installation
• Data Formats
• Configuration
• First Survey
• Quick Start