aimz 0.1.0__tar.gz

aimz-0.1.0/LICENSE

@@ -0,0 +1,201 @@
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   APPENDIX: How to apply the Apache License to your work.
+
+      To apply the Apache License to your work, attach the following
+      boilerplate notice, with the fields enclosed by brackets "[]"
+      replaced with your own identifying information. (Don't include
+      the brackets!)  The text should be enclosed in the appropriate
+      comment syntax for the file format. We also recommend that a
+      file or class name and description of purpose be included on the
+      same "printed page" as the copyright notice for easier
+      identification within third-party archives.
+
+   Copyright 2025 Eli Lilly and Company
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License.

aimz-0.1.0/PKG-INFO

@@ -0,0 +1,258 @@
+Metadata-Version: 2.4
+Name: aimz
+Version: 0.1.0
+Summary: Flexible probabilistic impact modeling at scale
+Author-email: Eunseop Kim <markean@pm.me>, Jirapat Samranvedhya <jirapat.samranvedhya@lilly.com>
+Maintainer-email: Eunseop Kim <markean@pm.me>
+License-Expression: Apache-2.0
+Project-URL: source, https://github.com/markean/aimz
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Science/Research
+Classifier: Topic :: Scientific/Engineering
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Requires-Python: >=3.11
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: arviz>=0.21
+Requires-Dist: dask>=2025.5
+Requires-Dist: jax>=0.5.3
+Requires-Dist: jax-dataloader>=0.1.3
+Requires-Dist: numpyro>=0.18.0
+Requires-Dist: scikit-learn>=1.6.1
+Requires-Dist: torch>=2.7
+Requires-Dist: tqdm>=4.67
+Requires-Dist: xarray>=2025.4
+Requires-Dist: zarr<4,>=3
+Provides-Extra: dev
+Requires-Dist: dill>=0.4.0; extra == "dev"
+Requires-Dist: pytest>=8.3; extra == "dev"
+Requires-Dist: pytest-cov>=6; extra == "dev"
+Provides-Extra: gpu
+Requires-Dist: jax[cuda12]>=0.5.3; extra == "gpu"
+Requires-Dist: torch>=2.7; extra == "gpu"
+Dynamic: license-file
+
+# aimz: Flexible probabilistic impact modeling at scale
+[![Python](https://img.shields.io/badge/python-3.11%20%7C%203.12%20%7C%203.13-blue.svg)]()
+[![Coverage Status](https://coveralls.io/repos/github/markean/aimz/badge.svg?branch=main)](https://coveralls.io/github/markean/aimz?branch=main)
+
+
+## Overview
+**aimz** is a Python library for flexible and scalable probabilistic impact modeling to assess the effects of interventions on outcomes of interest.
+Designed to work with user-defined models with probabilistic primitives, the library builds on [NumPyro](https://num.pyro.ai/en/stable/), [JAX](https://jax.readthedocs.io/en/latest/), [Xarray](https://xarray.dev/), and [Zarr](https://zarr.readthedocs.io/en/stable/) to enable efficient inference workflows.
+
+
+## Features
+- An intuitive API that combines ease of use from ML frameworks with the flexibility of probabilistic modeling.
+- Scalable computation via parallelism and distributed data processing—no manual orchestration required.
+- Variational inference as the primary inference engine, supporting custom optimization strategies and results.
+- Support for interventional causal inference for modeling counterfactuals and causal relations.
+
+
+## Installation
+CPU (default):
+```sh
+pip install -U aimz
+```
+
+GPU (NVIDIA, CUDA 12):
+```sh
+pip install -U "aimz[gpu]"
+```
+This installs `jax[cuda12]` with the version specified by the package. However, to ensure you have the latest compatible version of JAX with CUDA 12, it is recommended to update JAX separately after installation:
+```sh
+pip install -U "jax[cuda12]"
+```
+Refer to the [JAX installation guide](https://docs.jax.dev/en/latest/installation.html) for up-to-date compatibility and driver requirements.
+
+
+## Usage
+### Workflow
+1. Outline the model, considering the data generating process, latent variables, and causal relationships, if any.
+2. Translate the model into a **kernel** (i.e., a function) using NumPyro and JAX.
+3. Integrate the kernel into the provided API to train the model and perform inference.
+
+### Example 1: Regression Using a scikit-learn-like Workflow
+This example demonstrates a simple regression model following a typical ML workflow. The `ImpactModel` class provides `.fit()` for variational inference and posterior sampling, and `.predict()` for posterior predictive sampling. The optional `.cleanup()` removes posterior predictive samples saved as temporary files.
+```python
+import jax.numpy as jnp
+import numpy as np
+import numpyro.distributions as dist
+from jax import random
+from numpyro import optim, plate, sample
+from numpyro.infer import SVI, Trace_ELBO
+from numpyro.infer.autoguide import AutoNormal
+from sklearn.datasets import fetch_california_housing
+from sklearn.model_selection import train_test_split
+
+from aimz.model import ImpactModel
+
+# Load California Housing dataset
+housing = fetch_california_housing()
+X, y = housing.data, housing.target
+X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=42)
+
+
+# NumPyro model: linear regression
+def model(X: np.ndarray, y: np.ndarray | None = None) -> None:
+    """Bayesian linear regression."""
+    n_features = X.shape[1]
+
+    # Priors for weights, bias, and observation noise
+    w = sample("w", dist.Normal(jnp.zeros(n_features), jnp.ones(n_features)))
+    b = sample("b", dist.Normal())
+    sigma = sample("sigma", dist.Exponential())
+
+    # Plate over data
+    mu = jnp.dot(X, w) + b
+    with plate("data", X.shape[0]):
+        sample("y", dist.Normal(mu, sigma), obs=y)
+
+
+# Wrap with ImpactModel
+im = ImpactModel(
+    model,
+    rng_key=random.key(42),
+    vi=SVI(
+        model,
+        guide=AutoNormal(model),
+        optim=optim.Adam(step_size=1e-3),
+        loss=Trace_ELBO(),
+    ),
+)
+
+# Fit the model: variational inference followed by posterior sampling
+im.fit(X_train, y_train)
+
+# Predict on new data using posterior predictive sampling
+idata = im.predict(X_test)
+
+# Clean up posterior predictive samples saved to disk during `.predict()`
+im.cleanup()
+```
+> The `.fit()` step can be skipped if pre-trained variational inference results or posterior samples are available. These can be integrated into the `ImpactModel`, allowing `.predict()` to be available subsequently.
+
+### Example 2: Causal Network with Confounder
+This example illustrates a simple causal network. The variable `Z` has a direct causal effect on the outcome `Y`, while both are influenced by a shared confounder, `C`. An additional variable, `X`, is an observed exogenous factor that influences `Z` but has no direct effect on `Y`.
+
+Our objective is to estimate the causal effect of `Z` (or alternatively `X`) on `Y`, while properly accounting for the confounding influence of `C`. We assume the following generative model for the observed data:
+
+```python
+import jax
+import jax.numpy as jnp
+import numpyro.distributions as dist
+from jax import nn, random
+from numpyro import optim, plate, sample
+from numpyro.infer import SVI, Trace_ELBO, init_to_feasible
+from numpyro.infer.autoguide import AutoNormal
+
+from aimz.model import ImpactModel
+
+
+# NumPyro model: Z and y are influenced by C and X, with Z mediating part of y
+def model(X: jax.Array, C: jax.Array, y: jax.Array | None = None) -> None:
+    # Observed confounder
+    c = sample("c", dist.Exponential(), obs=C)
+
+    # Priors for coefficients in the structural model
+    # C -> Z and C -> Y
+    beta_cz = sample("beta_cz", dist.Normal())
+    beta_cy = sample("beta_cy", dist.Normal())
+
+    # X -> Z and Z -> Y
+    beta_xz = sample("beta_xz", dist.Normal())
+    beta_zy = sample("beta_zy", dist.Normal())
+
+    # Intercepts
+    beta_z = sample("beta_z", dist.Normal())
+    beta_y = sample("beta_y", dist.Normal())
+
+    # Observation noise for Z
+    sigma = sample("sigma", dist.Exponential())
+
+    # Plate over data
+    with plate("data", X.shape[0]):
+        mu_z = beta_z + beta_cz * c + beta_xz * X.squeeze(axis=1)
+        z = sample("z", dist.LogNormal(mu_z, sigma))
+
+        logits = beta_y + beta_cy * c + beta_zy * z
+        sample("y", dist.Bernoulli(logits=logits), obs=y)
+```
+
+#### Simulating data under a known structural model
+We generate synthetic data consistent with the assumed causal structure:
+- `C` is drawn from an exponential distribution.
+- `X` is a count variable from a Poisson distribution.
+- `Z` is generated as a noisy exponential function of `C` and `X`.
+- `Y` is a binary outcome influenced by both `C` and `Z` through a logistic model.
+
+```python
+# Create a pseudo-random number generator key for JAX
+rng_key = random.key(42)
+
+# Sample C from an Exponential distribution
+rng_key, rng_subkey = random.split(rng_key)
+C = random.exponential(rng_subkey, shape=(100,))
+
+# Sample X from a Poisson distribution
+rng_key, rng_subkey = random.split(rng_key)
+X = random.poisson(rng_subkey, lam=1, shape=(100, 1))
+
+# Generate Z influenced by C and X
+rng_key, rng_subkey = random.split(rng_key)
+mu_z = -1.0 + 0.5 * C - 1.5 * X.squeeze()
+sigma_z = 10.0  # Add substantial noise to reduce correlation between C and Z
+Z = jnp.exp(random.normal(rng_subkey, shape=(100,)) * sigma_z + mu_z)
+
+# Generate Y from a logistic regression on C and Z
+rng_key, rng_subkey = random.split(rng_key)
+logits = -2.0 + 5.0 * C + 0.1 * Z
+p = nn.sigmoid(logits)
+y = random.bernoulli(rng_subkey, p=p).astype(jnp.int32)
+```
+
+#### Fitting the model and estimating causal effects
+We fit the model using stochastic variational inference. Once trained, we perform a counterfactual analysis to isolate the effect of `Z` on `Y`.
+- `idata_factual` represents predictions under the factual setting (with observed `Z`).
+- `idata_counterfactual` represents predictions under a counterfactual intervention where `Z` is set to zero.
+Comparing these two distributions allows us to estimate the causal effect of `Z` on `Y`, adjusted for the influence of `C`.
+
+```python
+# Fit the model with SVI
+im = ImpactModel(
+    model,
+    rng_key=rng_key,
+    vi=SVI(
+        model,
+        guide=AutoNormal(model, init_loc_fn=init_to_feasible()),
+        optim=optim.Adam(step_size=1e-3),
+        loss=Trace_ELBO(),
+    ),
+)
+im.fit(X, y, C=C)
+
+# Predict under factual (Z) and counterfactual (zeroed Z) scenarios
+idata_factual = im.predict_on_batch(X, C=C, intervention={"z": Z})
+idata_counterfactual = im.predict_on_batch(
+    X,
+    C=C,
+    intervention={"z": jnp.zeros_like(Z)},
+)
+
+# Estimate causal effect of intervening on Z while conditioning on C
+impact = im.estimate_effect(
+    output_baseline=idata_factual,
+    output_intervention=idata_counterfactual,
+)
+```
+> Local latent variable requires `.predict_on_batch()` here. Prefer `.predict()` whenever it is compatible with the model.
+
+
+## Getting Help
+For feature requests, assistance, or any inquiries, contact maintainers or open an issue/pull request.
+
+
+## Contributing
+See [CONTRIBUTING.md](CONTRIBUTING.md) to get started.

aimz-0.1.0/README.md

@@ -0,0 +1,222 @@
+# aimz: Flexible probabilistic impact modeling at scale
+[![Python](https://img.shields.io/badge/python-3.11%20%7C%203.12%20%7C%203.13-blue.svg)]()
+[![Coverage Status](https://coveralls.io/repos/github/markean/aimz/badge.svg?branch=main)](https://coveralls.io/github/markean/aimz?branch=main)
+
+
+## Overview
+**aimz** is a Python library for flexible and scalable probabilistic impact modeling to assess the effects of interventions on outcomes of interest.
+Designed to work with user-defined models with probabilistic primitives, the library builds on [NumPyro](https://num.pyro.ai/en/stable/), [JAX](https://jax.readthedocs.io/en/latest/), [Xarray](https://xarray.dev/), and [Zarr](https://zarr.readthedocs.io/en/stable/) to enable efficient inference workflows.
+
+
+## Features
+- An intuitive API that combines ease of use from ML frameworks with the flexibility of probabilistic modeling.
+- Scalable computation via parallelism and distributed data processing—no manual orchestration required.
+- Variational inference as the primary inference engine, supporting custom optimization strategies and results.
+- Support for interventional causal inference for modeling counterfactuals and causal relations.
+
+
+## Installation
+CPU (default):
+```sh
+pip install -U aimz
+```
+
+GPU (NVIDIA, CUDA 12):
+```sh
+pip install -U "aimz[gpu]"
+```
+This installs `jax[cuda12]` with the version specified by the package. However, to ensure you have the latest compatible version of JAX with CUDA 12, it is recommended to update JAX separately after installation:
+```sh
+pip install -U "jax[cuda12]"
+```
+Refer to the [JAX installation guide](https://docs.jax.dev/en/latest/installation.html) for up-to-date compatibility and driver requirements.
+
+
+## Usage
+### Workflow
+1. Outline the model, considering the data generating process, latent variables, and causal relationships, if any.
+2. Translate the model into a **kernel** (i.e., a function) using NumPyro and JAX.
+3. Integrate the kernel into the provided API to train the model and perform inference.
+
+### Example 1: Regression Using a scikit-learn-like Workflow
+This example demonstrates a simple regression model following a typical ML workflow. The `ImpactModel` class provides `.fit()` for variational inference and posterior sampling, and `.predict()` for posterior predictive sampling. The optional `.cleanup()` removes posterior predictive samples saved as temporary files.
+```python
+import jax.numpy as jnp
+import numpy as np
+import numpyro.distributions as dist
+from jax import random
+from numpyro import optim, plate, sample
+from numpyro.infer import SVI, Trace_ELBO
+from numpyro.infer.autoguide import AutoNormal
+from sklearn.datasets import fetch_california_housing
+from sklearn.model_selection import train_test_split
+
+from aimz.model import ImpactModel
+
+# Load California Housing dataset
+housing = fetch_california_housing()
+X, y = housing.data, housing.target
+X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=42)
+
+
+# NumPyro model: linear regression
+def model(X: np.ndarray, y: np.ndarray | None = None) -> None:
+    """Bayesian linear regression."""
+    n_features = X.shape[1]
+
+    # Priors for weights, bias, and observation noise
+    w = sample("w", dist.Normal(jnp.zeros(n_features), jnp.ones(n_features)))
+    b = sample("b", dist.Normal())
+    sigma = sample("sigma", dist.Exponential())
+
+    # Plate over data
+    mu = jnp.dot(X, w) + b
+    with plate("data", X.shape[0]):
+        sample("y", dist.Normal(mu, sigma), obs=y)
+
+
+# Wrap with ImpactModel
+im = ImpactModel(
+    model,
+    rng_key=random.key(42),
+    vi=SVI(
+        model,
+        guide=AutoNormal(model),
+        optim=optim.Adam(step_size=1e-3),
+        loss=Trace_ELBO(),
+    ),
+)
+
+# Fit the model: variational inference followed by posterior sampling
+im.fit(X_train, y_train)
+
+# Predict on new data using posterior predictive sampling
+idata = im.predict(X_test)
+
+# Clean up posterior predictive samples saved to disk during `.predict()`
+im.cleanup()
+```
+> The `.fit()` step can be skipped if pre-trained variational inference results or posterior samples are available. These can be integrated into the `ImpactModel`, allowing `.predict()` to be available subsequently.
+
+### Example 2: Causal Network with Confounder
+This example illustrates a simple causal network. The variable `Z` has a direct causal effect on the outcome `Y`, while both are influenced by a shared confounder, `C`. An additional variable, `X`, is an observed exogenous factor that influences `Z` but has no direct effect on `Y`.
+
+Our objective is to estimate the causal effect of `Z` (or alternatively `X`) on `Y`, while properly accounting for the confounding influence of `C`. We assume the following generative model for the observed data:
+
+```python
+import jax
+import jax.numpy as jnp
+import numpyro.distributions as dist
+from jax import nn, random
+from numpyro import optim, plate, sample
+from numpyro.infer import SVI, Trace_ELBO, init_to_feasible
+from numpyro.infer.autoguide import AutoNormal
+
+from aimz.model import ImpactModel
+
+
+# NumPyro model: Z and y are influenced by C and X, with Z mediating part of y
+def model(X: jax.Array, C: jax.Array, y: jax.Array | None = None) -> None:
+    # Observed confounder
+    c = sample("c", dist.Exponential(), obs=C)
+
+    # Priors for coefficients in the structural model
+    # C -> Z and C -> Y
+    beta_cz = sample("beta_cz", dist.Normal())
+    beta_cy = sample("beta_cy", dist.Normal())
+
+    # X -> Z and Z -> Y
+    beta_xz = sample("beta_xz", dist.Normal())
+    beta_zy = sample("beta_zy", dist.Normal())
+
+    # Intercepts
+    beta_z = sample("beta_z", dist.Normal())
+    beta_y = sample("beta_y", dist.Normal())
+
+    # Observation noise for Z
+    sigma = sample("sigma", dist.Exponential())
+
+    # Plate over data
+    with plate("data", X.shape[0]):
+        mu_z = beta_z + beta_cz * c + beta_xz * X.squeeze(axis=1)
+        z = sample("z", dist.LogNormal(mu_z, sigma))
+
+        logits = beta_y + beta_cy * c + beta_zy * z
+        sample("y", dist.Bernoulli(logits=logits), obs=y)
+```
+
+#### Simulating data under a known structural model
+We generate synthetic data consistent with the assumed causal structure:
+- `C` is drawn from an exponential distribution.
+- `X` is a count variable from a Poisson distribution.
+- `Z` is generated as a noisy exponential function of `C` and `X`.
+- `Y` is a binary outcome influenced by both `C` and `Z` through a logistic model.
+
+```python
+# Create a pseudo-random number generator key for JAX
+rng_key = random.key(42)
+
+# Sample C from an Exponential distribution
+rng_key, rng_subkey = random.split(rng_key)
+C = random.exponential(rng_subkey, shape=(100,))
+
+# Sample X from a Poisson distribution
+rng_key, rng_subkey = random.split(rng_key)
+X = random.poisson(rng_subkey, lam=1, shape=(100, 1))
+
+# Generate Z influenced by C and X
+rng_key, rng_subkey = random.split(rng_key)
+mu_z = -1.0 + 0.5 * C - 1.5 * X.squeeze()
+sigma_z = 10.0  # Add substantial noise to reduce correlation between C and Z
+Z = jnp.exp(random.normal(rng_subkey, shape=(100,)) * sigma_z + mu_z)
+
+# Generate Y from a logistic regression on C and Z
+rng_key, rng_subkey = random.split(rng_key)
+logits = -2.0 + 5.0 * C + 0.1 * Z
+p = nn.sigmoid(logits)
+y = random.bernoulli(rng_subkey, p=p).astype(jnp.int32)
+```
+
+#### Fitting the model and estimating causal effects
+We fit the model using stochastic variational inference. Once trained, we perform a counterfactual analysis to isolate the effect of `Z` on `Y`.
+- `idata_factual` represents predictions under the factual setting (with observed `Z`).
+- `idata_counterfactual` represents predictions under a counterfactual intervention where `Z` is set to zero.
+Comparing these two distributions allows us to estimate the causal effect of `Z` on `Y`, adjusted for the influence of `C`.
+
+```python
+# Fit the model with SVI
+im = ImpactModel(
+    model,
+    rng_key=rng_key,
+    vi=SVI(
+        model,
+        guide=AutoNormal(model, init_loc_fn=init_to_feasible()),
+        optim=optim.Adam(step_size=1e-3),
+        loss=Trace_ELBO(),
+    ),
+)
+im.fit(X, y, C=C)
+
+# Predict under factual (Z) and counterfactual (zeroed Z) scenarios
+idata_factual = im.predict_on_batch(X, C=C, intervention={"z": Z})
+idata_counterfactual = im.predict_on_batch(
+    X,
+    C=C,
+    intervention={"z": jnp.zeros_like(Z)},
+)
+
+# Estimate causal effect of intervening on Z while conditioning on C
+impact = im.estimate_effect(
+    output_baseline=idata_factual,
+    output_intervention=idata_counterfactual,
+)
+```
+> Local latent variable requires `.predict_on_batch()` here. Prefer `.predict()` whenever it is compatible with the model.
+
+
+## Getting Help
+For feature requests, assistance, or any inquiries, contact maintainers or open an issue/pull request.
+
+
+## Contributing
+See [CONTRIBUTING.md](CONTRIBUTING.md) to get started.