.. This file is generated by scripts/render_dataset_docs.py. Do not edit by hand.

FPA-FOD Tabular
===============

Incident-level FPA-FOD features packaged for wildfire cause and size classification.

Overview
--------

FPA-FOD Tabular converts one wildfire incident record into one feature vector for classification tasks such as incident cause prediction and grouped size prediction.

PyHazards exposes it as a loadable dataset with a deterministic micro mode so the full source database is not required for smoke tests or quick experimentation.

At a Glance
-----------

.. list-table::
   :widths: 28 72
   :stub-columns: 1

   * - Provider
     - Fire Program Analysis Fire-Occurrence Database (FPA-FOD) adaptation in PyHazards
   * - Hazard Family
     - Wildfire
   * - Source Role
     - Incident Tabular
   * - Coverage
     - User-provided FPA-FOD coverage
   * - Geometry
     - Tabular feature vectors
   * - Spatial Resolution
     - Incident-level records
   * - Temporal Resolution
     - Event-based
   * - Update Cadence
     - User-managed local inputs or deterministic micro mode
   * - Period of Record
     - Depends on the supplied FPA-FOD source files
   * - Formats
     - SQLite, DB, CSV, and Parquet inputs
   * - Inspection CLI
     - ``python -m pyhazards.datasets.fpa_fod_tabular.inspection --task cause --micro``
   * - Registry Entry
     - ``fpa_fod_tabular``

Data Characteristics
--------------------

- Supports task='cause' and task='size' classification targets.
- Accepts SQLite, DB, CSV, and Parquet sources.
- Micro mode keeps the path deterministic and lightweight for validation.
- Returned splits follow the standard DataBundle contract with tabular inputs and integer targets.

Typical Use Cases
~~~~~~~~~~~~~~~~~

- Wildfire cause classification experiments.
- Grouped fire size classification from incident records.
- Lightweight smoke and regression tests for the wildfire tabular path.

Access
------

Use the links below to access the upstream source or its public documentation.

- `PyHazards public dataset catalog </pyhazards_datasets.html>`_

PyHazards Usage
---------------

Use this dataset through the public inspection or registry surface documented below.

Registry Workflow
~~~~~~~~~~~~~~~~~

Primary dataset name: ``fpa_fod_tabular``

.. code-block:: python

   from pyhazards.datasets import load_dataset

   data = load_dataset(
       "fpa_fod_tabular",
       task="cause",
       micro=True,
       normalize=True,
   ).load()

   train = data.get_split("train")
   print(train.inputs.shape, train.targets.shape)

- region='US' uses all available states, while region='CA' restricts to California incidents.
- cause_mode='paper5' preserves the five consolidated cause groups used by the public wildfire tabular path.

Related Coverage
~~~~~~~~~~~~~~~~

**Benchmarks:** :doc:`Wildfire Benchmark </benchmarks/wildfire_benchmark>`

**Representative Models:** :doc:`DNN-LSTM-AutoEncoder </modules/models_wildfire_fpa>`

Inspection Workflow
-------------------

Use the documented inspection path below to validate local files before training or analysis.

.. code-block:: bash

   python -m pyhazards.datasets.fpa_fod_tabular.inspection --task cause --micro

Reference
---------

- `PyHazards FPA-FOD tabular adaptation for the wildfire incident classification path. <https://github.com/LabRAI/PyHazards>`_.