From 1fc85be50dc143fc289546b04d841487afced49f Mon Sep 17 00:00:00 2001
From: "copilot-swe-agent[bot]" <198982749+Copilot@users.noreply.github.com>
Date: Thu, 27 Nov 2025 16:01:24 +0000
Subject: [PATCH 1/3] Initial plan


From d4991ac0ef1bb45cb8613b276ecffb8ea6f6712f Mon Sep 17 00:00:00 2001
From: "copilot-swe-agent[bot]" <198982749+Copilot@users.noreply.github.com>
Date: Thu, 27 Nov 2025 16:05:24 +0000
Subject: [PATCH 2/3] Add Copilot instructions for PyMC-Marketing repository

Co-authored-by: williambdean <57733339+williambdean@users.noreply.github.com>
---
 .github/copilot-instructions.md | 188 ++++++++++++++++++++++++++++++++
 1 file changed, 188 insertions(+)
 create mode 100644 .github/copilot-instructions.md

diff --git a/.github/copilot-instructions.md b/.github/copilot-instructions.md
new file mode 100644
index 000000000..b1d74b2b1
--- /dev/null
+++ b/.github/copilot-instructions.md
@@ -0,0 +1,188 @@
+# Copilot Instructions for PyMC-Marketing
+
+This document provides instructions for GitHub Copilot when working in the PyMC-Marketing repository.
+
+## Project Overview
+
+PyMC-Marketing is a Python library for Bayesian marketing analytics built on top of PyMC. The library provides tools for:
+
+- **Marketing Mix Modeling (MMM)**: Bayesian models for measuring marketing effectiveness, including adstock transformations, saturation effects, and budget optimization
+- **Customer Lifetime Value (CLV)**: Models for predicting customer value, including BG/NBD, Pareto/NBD, and Gamma-Gamma models
+- **Customer Choice Analysis**: Multivariate Interrupted Time Series (MVITS) models for analyzing product launches
+- **Bass Diffusion Models**: For predicting adoption of new products
+
+## Project Structure
+
+```
+pymc_marketing/
+├── mmm/           # Marketing Mix Modeling components
+├── clv/           # Customer Lifetime Value models
+├── customer_choice/  # Customer choice analysis (MVITS)
+├── bass/          # Bass diffusion models
+├── model_builder.py  # Base model building utilities
+└── prior.py       # Prior distribution utilities
+```
+
+## Development Setup
+
+### Environment Setup
+
+```bash
+# Create conda environment
+conda env create -f environment.yml
+conda activate pymc-marketing-dev
+
+# Install package in editable mode with development dependencies
+make init
+```
+
+### Pre-commit Hooks
+
+```bash
+pre-commit install
+```
+
+## Code Style and Conventions
+
+### Formatting and Linting
+
+- **Formatter**: Ruff (configured in `pyproject.toml`)
+- **Linter**: Ruff and mypy
+- **Max line length**: 120 characters
+- **Docstring format**: NumPy style
+
+```bash
+# Check linting
+make check_lint
+
+# Auto-fix linting issues
+make lint
+
+# Check formatting
+make check_format
+
+# Auto-fix formatting
+make format
+```
+
+### Documentation Style
+
+All public methods must have informative docstrings following the [NumPy docstring format](https://numpydoc.readthedocs.io/en/latest/format.html):
+
+```python
+def function_name(param1: type, param2: type) -> return_type:
+    """Short description of the function.
+
+    Longer description if needed.
+
+    Parameters
+    ----------
+    param1 : type
+        Description of param1.
+    param2 : type
+        Description of param2.
+
+    Returns
+    -------
+    return_type
+        Description of return value.
+
+    Examples
+    --------
+    >>> function_name(value1, value2)
+    expected_output
+    """
+```
+
+### Type Hints
+
+- Use type hints for all function parameters and return values
+- Import types from `typing` module when needed
+- The codebase uses `py.typed` marker for PEP 561 compliance
+
+### Testing
+
+```bash
+# Run all tests
+make test
+
+# Run specific test file
+pytest tests/path/to/test_file.py
+
+# Run with verbose output
+pytest -v tests/
+```
+
+Tests are located in the `tests/` directory and mirror the structure of `pymc_marketing/`.
+
+## Key Dependencies
+
+- **PyMC** (>=5.24.1): Core probabilistic programming framework
+- **PyTensor** (>=2.31.3): Tensor computation library
+- **ArviZ** (>=0.13.0): Bayesian model diagnostics and visualization
+- **pandas**: Data manipulation
+- **NumPy**: Numerical computing
+- **scikit-learn** (>=1.1.1): Machine learning utilities
+- **xarray**: Multi-dimensional arrays with labels
+- **Pydantic** (>=2.1.0): Data validation
+
+## Common Patterns
+
+### Model Building
+
+Models in PyMC-Marketing typically inherit from base classes in `model_builder.py`:
+
+```python
+from pymc_marketing.model_builder import ModelBuilder
+
+class MyModel(ModelBuilder):
+    # Model implementation
+    pass
+```
+
+### Prior Distributions
+
+Use the `Prior` class from `pymc_marketing.prior` for defining priors:
+
+```python
+from pymc_marketing.prior import Prior
+
+prior = Prior("Normal", mu=0, sigma=1)
+```
+
+### Data Loading
+
+Use paths from `pymc_marketing.paths` for accessing example data:
+
+```python
+from pymc_marketing.paths import data_dir
+
+file_path = data_dir / "example_data.csv"
+```
+
+## Building Documentation
+
+```bash
+# Build HTML documentation
+make html
+
+# Clean and rebuild
+make cleandocs
+make html
+```
+
+Documentation is in `docs/source/` and built with Sphinx.
+
+## Pull Request Guidelines
+
+- Reference the related issue in the PR description
+- Include tests for new functionality
+- Update documentation for API changes
+- Ensure all linting and tests pass
+- Follow the PR checklist in `.github/pull_request_template.md`
+
+## Additional Resources
+
+- [CONTRIBUTING.md](../CONTRIBUTING.md): Full contribution guidelines
+- [PyMC-Marketing Documentation](https://www.pymc-marketing.io)
+- [PyMC Documentation](https://www.pymc.io)

From 8172a6734777dc21d2238409c3a992898857def0 Mon Sep 17 00:00:00 2001
From: "copilot-swe-agent[bot]" <198982749+Copilot@users.noreply.github.com>
Date: Thu, 27 Nov 2025 16:58:52 +0000
Subject: [PATCH 3/3] Address PR review feedback on copilot instructions

Co-authored-by: williambdean <57733339+williambdean@users.noreply.github.com>
---
 .github/copilot-instructions.md | 31 ++++++++++++++++++++++++-------
 1 file changed, 24 insertions(+), 7 deletions(-)

diff --git a/.github/copilot-instructions.md b/.github/copilot-instructions.md
index b1d74b2b1..470db7227 100644
--- a/.github/copilot-instructions.md
+++ b/.github/copilot-instructions.md
@@ -89,8 +89,17 @@ def function_name(param1: type, param2: type) -> return_type:
 
     Examples
     --------
+    Doctest style:
+
     >>> function_name(value1, value2)
     expected_output
+
+    Or Sphinx code block style:
+
+    .. code-block:: python
+
+        result = function_name(value1, value2)
+
     """
 ```
 
@@ -102,7 +111,12 @@ def function_name(param1: type, param2: type) -> return_type:
 
 ### Testing
 
+Ensure the conda environment is activated before running tests:
+
 ```bash
+# Activate environment first
+conda activate pymc-marketing-dev
+
 # Run all tests
 make test
 
@@ -117,14 +131,17 @@ Tests are located in the `tests/` directory and mirror the structure of `pymc_ma
 
 ## Key Dependencies
 
-- **PyMC** (>=5.24.1): Core probabilistic programming framework
-- **PyTensor** (>=2.31.3): Tensor computation library
-- **ArviZ** (>=0.13.0): Bayesian model diagnostics and visualization
+See `pyproject.toml` for version requirements. Main dependencies include:
+
+- **PyMC**: Core probabilistic programming framework
+- **PyTensor**: Tensor computation library
+- **ArviZ**: Bayesian model diagnostics and visualization
 - **pandas**: Data manipulation
 - **NumPy**: Numerical computing
-- **scikit-learn** (>=1.1.1): Machine learning utilities
+- **scikit-learn**: Machine learning utilities
 - **xarray**: Multi-dimensional arrays with labels
-- **Pydantic** (>=2.1.0): Data validation
+- **Pydantic**: Data validation
+- **pymc-extras**: Extended PyMC utilities (priors, etc.)
 
 ## Common Patterns
 
@@ -142,10 +159,10 @@ class MyModel(ModelBuilder):
 
 ### Prior Distributions
 
-Use the `Prior` class from `pymc_marketing.prior` for defining priors:
+Use the `Prior` class from `pymc_extras.prior` for defining priors:
 
 ```python
-from pymc_marketing.prior import Prior
+from pymc_extras.prior import Prior
 
 prior = Prior("Normal", mu=0, sigma=1)
 ```