Merge pull request #299 from Jammy2211/docs/quantity

docs: add module docstrings to quantity package

Author: Jammy2211

autogalaxy/quantity/dataset_quantity.py

@@ -1,3 +1,13 @@
+"""
+`DatasetQuantity` — a dataset wrapper for derived lensing quantities.
+
+This module provides `DatasetQuantity`, which wraps an arbitrary array-valued quantity (e.g. convergence,
+deflection angles, surface brightness) together with a noise-map, so that the quantity can be treated as
+a dataset and fitted via `FitQuantity` and `AnalysisQuantity`.
+
+The noise-map enables the computation of chi-squared values and log-likelihoods, turning a pure
+comparison of model quantities into a statistically principled goodness-of-fit metric.
+"""
 import logging
 import numpy as np
 from pathlib import Path

autogalaxy/quantity/fit_quantity.py

@@ -1,3 +1,19 @@
+"""
+Fit a `DatasetQuantity` object with model quantities computed from light or mass profiles.
+
+`FitQuantity` provides a general-purpose fitting framework for derived lensing quantities (e.g. convergence,
+deflection angles, surface brightness). Given a `DatasetQuantity` (which contains a target quantity array and
+a noise-map), `FitQuantity` evaluates the same quantity from a model object and computes the likelihood of
+the model matching the target.
+
+This is useful for, for example:
+
+- Matching the convergence of a power-law model to the convergence of an NFW profile.
+- Comparing deflection angles from two different mass distributions.
+- Fitting a pixelized reconstruction to a separately computed model quantity.
+
+The `AnalysisQuantity` class uses `FitQuantity` internally to enable non-linear fitting of these quantities.
+"""
 from typing import List, Optional, Union
 
 import autoarray as aa