Skip to main content

Generative materials benchmarking metrics, inspired by CDVAE.

Project description

Project generated with PyScaffold ReadTheDocs Coveralls PyPI-Server Conda-Forge PyPI - Downloads Lines of code

This is not an official repository of Matbench, but eventually, it may be incorporated into Matbench

matbench-genmetrics Open In Colab

Generative materials benchmarking metrics, inspired by guacamol and CDVAE.

This repository provides standardized benchmarks for benchmarking generative models for crystal structure. Each benchmark has a fixed dataset, a predefined split, and a notion of best (i.e. metric) associated with it.

Getting Started

Installation, a dummy example, output metrics for the example, and descriptions of the benchmark metrics.


Create a conda environment with the matbench-genmetrics package installed from the conda-forge channel. Then activate the environment.

**NOTE: not available on conda-forge as of 2022-07-30, recipe under review by conda-forge team. So use pip install matbench-genmetrics for now

conda create --name matbench-genmetrics --channel conda-forge python==3.9.* matbench-genmetrics
conda activate matbench-genmetrics

NOTE: It doesn't have to be Python 3.9; you can remove python==3.9.* altogether or change this to e.g. python==3.8.*. See Advanced Installation


NOTE: be sure to set dummy=False for the real/full benchmark run. MPTSMetrics10 is intended for fast prototyping and debugging, as it assumes only 10 generated structures.

>>> from tqdm import tqdm
>>> from mp_time_split.utils.gen import DummyGenerator
>>> from matbench_genmetrics.core import MPTSMetrics10, MPTSMetrics100, MPTSMetrics1000, MPTSMetrics10000
>>> mptm = MPTSMetrics10(dummy=True)
>>> for fold in mptm.folds:
>>>     train_val_inputs = mptm.get_train_and_val_data(fold)
>>>     dg = DummyGenerator()
>>>     gen_structures = dg.gen(n=mptm.num_gen)
>>>     mptm.evaluate_and_record(fold, gen_structures)


    0: {
        "validity": 0.4375,
        "coverage": 0.0,
        "novelty": 1.0,
        "uniqueness": 0.9777777777777777,
    1: {
        "validity": 0.4390681003584229,
        "coverage": 0.0,
        "novelty": 1.0,
        "uniqueness": 0.9333333333333333,
    2: {
        "validity": 0.4401197604790419,
        "coverage": 0.0,
        "novelty": 1.0,
        "uniqueness": 0.8222222222222222,
    3: {
        "validity": 0.4408740359897172,
        "coverage": 0.0,
        "novelty": 1.0,
        "uniqueness": 0.8444444444444444,
    4: {
        "validity": 0.4414414414414415,
        "coverage": 0.0,
        "novelty": 1.0,
        "uniqueness": 0.9111111111111111,


Metric Description
Validity One minus (Wasserstein distance between distribution of space group numbers for train and generated structures divided by distance of dummy case between train and space_group_number == 1). See also
Coverage Match counts between held-out test structures and generated structures divided by number of test structures ("predict the future").
Novelty One minus (match counts between train structures and generated structures divided by number of generated structures).
Uniqueness One minus (non-self-comparing match counts within generated structures divided by total possible non-self-comparing matches).

A match is when StructureMatcher(stol=0.5, ltol=0.3, angle_tol=10.0).fit(s1, s2) evaluates to True.

Detailed descriptions of the metrics are given in

Advanced Installation

Anaconda (conda) installation (recommended)

(2022-07-30, conda-forge installation pending, fallback to pip install matbench-genmetrics as separate command)

Create and activate a new conda environment named matbench-genmetrics (-n) that will search for and install the matbench-genmetrics package from the conda-forge Anaconda channel (-c).

conda env create -n matbench-genmetrics -c conda-forge matbench-genmetrics
conda activate matbench-genmetrics

Alternatively, in an already activated environment:

conda install -c conda-forge matbench-genmetrics

If you run into conflicts with packages you are integrating with matbench-genmetrics, please try installing all packages in a single line of code (or two if mixing conda and pip packages in the same environment) and installing with mamba (source).

PyPI (pip) installation

Create and activate a new conda environment named matbench-genmetrics (-n) with python==3.9.* or your preferred Python version, then install matbench-genmetrics via pip.

conda create -n matbench-genmetrics python==3.9.*
conda activate matbench-genmetrics
pip install matbench-genmetrics

Editable installation

In order to set up the necessary environment:

  1. clone and enter the repository via:

    git clone
    cd matbench-genmetrics
  2. create and activate a new conda environment (optional, but recommended)

    conda env create --name matbench-genmetrics python==3.9.*
    conda activate matbench-genmetrics
  3. perform an editable (-e) installation in the current directory (.):

    pip install -e .

NOTE: Some changes, e.g. in setup.cfg, might require you to run pip install -e . again.

Optional and needed only once after git clone:

  1. install several pre-commit git hooks with:

    pre-commit install
    # You might also want to run `pre-commit autoupdate`

    and checkout the configuration under .pre-commit-config.yaml. The -n, --no-verify flag of git commit can be used to deactivate pre-commit hooks temporarily.

  2. install nbstripout git hooks to remove the output cells of committed notebooks with:

    nbstripout --install --attributes notebooks/.gitattributes

    This is useful to avoid large diffs due to plots in your notebooks. A simple nbstripout --uninstall will revert these changes.

Then take a look into the scripts and notebooks folders.

Dependency Management & Reproducibility

  1. Always keep your abstract (unpinned) dependencies updated in environment.yml and eventually in setup.cfg if you want to ship and install your package via pip later on.

  2. Create concrete dependencies as environment.lock.yml for the exact reproduction of your environment with:

    conda env export -n matbench-genmetrics -f environment.lock.yml

    For multi-OS development, consider using --no-builds during the export.

  3. Update your current environment with respect to a new environment.lock.yml using:

    conda env update -f environment.lock.yml --prune

Project Organization

├──              <- List of developers and maintainers.
├──            <- Changelog to keep track of new features and fixes.
├──         <- Guidelines for contributing to this project.
├── Dockerfile              <- Build a docker container with `docker build .`.
├── LICENSE.txt             <- License as chosen on the command-line.
├──               <- The top-level README for developers.
├── configs                 <- Directory for configurations of model & application.
├── data
│   ├── external            <- Data from third party sources.
│   ├── interim             <- Intermediate data that has been transformed.
│   ├── processed           <- The final, canonical data sets for modeling.
│   └── raw                 <- The original, immutable data dump.
├── docs                    <- Directory for Sphinx documentation in rst or md.
├── environment.yml         <- The conda environment file for reproducibility.
├── models                  <- Trained and serialized models, model predictions,
│                              or model summaries.
├── notebooks               <- Jupyter notebooks. Naming convention is a number (for
│                              ordering), the creator's initials and a description,
│                              e.g. `1.0-fw-initial-data-exploration`.
├── pyproject.toml          <- Build configuration. Don't change! Use `pip install -e .`
│                              to install for development or to build `tox -e build`.
├── references              <- Data dictionaries, manuals, and all other materials.
├── reports                 <- Generated analysis as HTML, PDF, LaTeX, etc.
│   └── figures             <- Generated plots and figures for reports.
├── scripts                 <- Analysis and production scripts which import the
│                              actual PYTHON_PKG, e.g. train_model.
├── setup.cfg               <- Declarative configuration of your project.
├──                <- [DEPRECATED] Use `python develop` to install for
│                              development or `python bdist_wheel` to build.
├── src
│   └── matbench_genmetrics <- Actual Python package where the main functionality goes.
├── tests                   <- Unit tests which can be run with `pytest`.
├── .coveragerc             <- Configuration for coverage reports of unit tests.
├── .isort.cfg              <- Configuration for git hook that sorts imports.
└── .pre-commit-config.yaml <- Configuration of pre-commit git hooks.


This project has been set up using PyScaffold 4.2.2.post1.dev2+ge50b5e1 and the dsproject extension 0.7.2.post1.dev2+geb5d6b6.

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

matbench-genmetrics-0.6.1.tar.gz (4.6 MB view hashes)

Uploaded source

Built Distribution

matbench_genmetrics-0.6.1-py3-none-any.whl (51.6 kB view hashes)

Uploaded py3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page