Data

Tools

Indexes

Applications

Experiments

Open Source Science

Recent Releases of https://github.com/ppdebreuck/modnet

https://github.com/ppdebreuck/modnet - v0.4.4

What's Changed

  • Usability tweaks & new featurizer preset by @ml-evs in https://github.com/ppdebreuck/modnet/pull/215

Full Changelog: https://github.com/ppdebreuck/modnet/compare/v0.4.3...v0.4.4

- Jupyter Notebook
Published by ppdebreuck about 2 years ago

https://github.com/ppdebreuck/modnet - v0.4.3

What's Changed

  • fixed bugged conditionals in evaluate() by @kyledmiller in https://github.com/ppdebreuck/modnet/pull/210
  • Add simple test for evaluate by @ml-evs in https://github.com/ppdebreuck/modnet/pull/211

New Contributors

  • @kyledmiller made their first contribution in https://github.com/ppdebreuck/modnet/pull/210

Full Changelog: https://github.com/ppdebreuck/modnet/compare/v0.4.2...v0.4.3

- Jupyter Notebook
Published by ppdebreuck over 2 years ago

https://github.com/ppdebreuck/modnet - v0.4.2

What's Changed

  • Deprecated BayesianMODNetModel and update deps by @ml-evs in https://github.com/ppdebreuck/modnet/pull/182
  • Fix issue with fit_preset invoking fit incorrectly during refit by @ml-evs in https://github.com/ppdebreuck/modnet/pull/181
  • 3.10 compatibility by @ppdebreuck in https://github.com/ppdebreuck/modnet/pull/198
  • Improve evaluate (custom loss, ...) by @ppdebreuck in https://github.com/ppdebreuck/modnet/pull/194
  • Drop Python 3.8 and update other deps by @ml-evs in https://github.com/ppdebreuck/modnet/pull/201
  • Bump matminer version by @ml-evs in https://github.com/ppdebreuck/modnet/pull/199
  • Attempt at bumping pymatgen and matminer by @ml-evs in https://github.com/ppdebreuck/modnet/pull/203
  • Backwards compatibility of test data with pymatgen by @ml-evs in https://github.com/ppdebreuck/modnet/pull/206
  • Properly handle Bayesian model import failure by @ml-evs in https://github.com/ppdebreuck/modnet/pull/207

Full Changelog: https://github.com/ppdebreuck/modnet/compare/v0.4.1...v0.4.2

- Jupyter Notebook
Published by ppdebreuck over 2 years ago

https://github.com/ppdebreuck/modnet - v0.4.1

What's Changed

  • Fixed refit=0 in FitGenetic, it behaves as before (ensemble of 10 best architecture, ensembled over the nested(default 5) folds)
  • Bump pymatgen from 2023.1.30 to 2023.7.20, compatible with cython 3

- Jupyter Notebook
Published by ppdebreuck almost 3 years ago

https://github.com/ppdebreuck/modnet - v0.4.0

What's Changed

  • /!\ New default model architecture v0.4.0 changes the default architecture of all MODNet models. It is now possible to predict vectors (previously one had to make individual joint learned properties - which can be slow when the output dimensionality is high), while keeping the joint-learning architecture. In essence, the architecture is moving to joint-learning on vectors. Previously saved models are still compatible and will be loaded following the old architecture. Please consider retraining your saved models in the near future as modnet will transition to v1.0 without support of the old model architecture. See https://github.com/ppdebreuck/modnet/pull/89 and https://github.com/ppdebreuck/modnet/pull/155 by @ppdebreuck

  • Possibility to remove or not fully NaNs features by @gbrunin in https://github.com/ppdebreuck/modnet/pull/157

Full Changelog: https://github.com/ppdebreuck/modnet/compare/v0.3.1...v0.4.0

- Jupyter Notebook
Published by ppdebreuck almost 3 years ago

https://github.com/ppdebreuck/modnet - v0.3.1

What's Changed

  • Fix backward compatibility. Models <v0.3 can safely be loaded. By @ppdebreuck in https://github.com/ppdebreuck/modnet/pull/153
  • Tweak README by @ml-evs in https://github.com/ppdebreuck/modnet/pull/154

Full Changelog: https://github.com/ppdebreuck/modnet/compare/v0.3.0...v0.3.1

- Jupyter Notebook
Published by ppdebreuck almost 3 years ago

https://github.com/ppdebreuck/modnet - v0.3.0

What's Changed

1) Impute missing values by @gbrunin in https://github.com/ppdebreuck/modnet/pull/149 After the featurization, the NaNs are not replaced by 0 anymore. The infinite values are replaced by NaNs. Then, the NaNs are handled when fitting the model using a SimpleImputer which can be chosen. It is then stored as an attribute to the model, and can be re-used when predicting new values. The scaler can also be chosen (StandardScaler or MinMaxScaler), and the user can also choose to first impute then scale, or first scale then impute. Both can be argued (do we want to keep the same distribution as the initial feature, or to change it by moving the NaNs outside the distribution).

2) New featurizer presets by @gbrunin in https://github.com/ppdebreuck/modnet/pull/150 The full list of featurizers are: - DeBreuck2020Featurizer, - CompositionOnlyFeaturizer, - Matminer2023Featurizer, - MatminerAll2023Featurizer, - CompositionOnlyMatminer2023Featurizer, - CompositionOnlyMatminerAll2023Featurizer, It also adds the possibility to use only features that are continuous with respect to the composition. Some features are by their nature not continuous, which can lead to unphysical discontinuities when predicting a property as a function of the materials composition.

3) Introducing better customisation by @ppdebreuck in https://github.com/ppdebreuck/modnet/pull/148 - Running feature selection only on a subset of properties present in the MODData. feature_ selection() now enables this with ignore_names. - By default, FitGenetic will proceed by using joint-learning when multiple targets are given in the MODData. This can now be avoided by using ignore_names in FitGenetic(). - MODNetModel.fit() can take optional fit_params that are passed through to Keras model.fit(). - fit_params can also be passed to FitGenetic.run() - MODNetModel.fit() can take a custom loss function. - FitGenetic() can take a custom loss function. - Custom data can be passed trough MODNetModel.fit(). It will be appended to the targets (axis=-1). This can be useful for defining custom loss functions. - Any property called custom_data in FitGenetic is ignored, and appended to the targets (axis=-1). This can be useful for defining custom loss functions.

4) Add getparams and setparams to the MODNet model by @gbrunin in https://github.com/ppdebreuck/modnet/pull/151 This includes renaming of the the EnsembleMODNetModel ´modnet_models´ arg to ´models´

New Contributors 🎉

  • @gbrunin made their first contribution in https://github.com/ppdebreuck/modnet/pull/149 Thanks Guillaume Brunin !

Full Changelog: https://github.com/ppdebreuck/modnet/compare/v0.2.1...v0.3.0

- Jupyter Notebook
Published by ppdebreuck about 3 years ago

https://github.com/ppdebreuck/modnet - v0.2.1

What's Changed

  • Add support for Python 3.9 and 3.10 by @ml-evs in https://github.com/ppdebreuck/modnet/pull/119
  • Update dependency pins, compatibility with TF 2.11 and add install notes by @ml-evs in https://github.com/ppdebreuck/modnet/pull/122
  • Transfer some info from README into docs by @ml-evs in https://github.com/ppdebreuck/modnet/pull/136

Full Changelog: https://github.com/ppdebreuck/modnet/compare/v0.2.0...v0.2.1

- Jupyter Notebook
Published by ppdebreuck over 3 years ago

https://github.com/ppdebreuck/modnet - v0.2.0

What's Changed

  • Add new default feature preset and updates for new matminer & pymatgen versions by @ml-evs in https://github.com/ppdebreuck/modnet/pull/101
  • Bump tensorflow from 2.10.0 to 2.10.1 by @dependabot in https://github.com/ppdebreuck/modnet/pull/112
  • fix verbosity by @ppdebreuck in https://github.com/ppdebreuck/modnet/pull/128
  • Replace deprecated NumPy and Tensorflow calls by @ml-evs in https://github.com/ppdebreuck/modnet/pull/123
  • Add mode where each featurizer is applied individually by @ml-evs in https://github.com/ppdebreuck/modnet/pull/127

Full Changelog: https://github.com/ppdebreuck/modnet/compare/v0.1.13...v0.2.0

- Jupyter Notebook
Published by ppdebreuck over 3 years ago

https://github.com/ppdebreuck/modnet - v0.1.13

What's Changed

  • Add pinned requirements file by @ml-evs in https://github.com/ppdebreuck/modnet/pull/94
  • Make sure new deps do not get overwritten by CI by @ml-evs in https://github.com/ppdebreuck/modnet/pull/99
  • Add instructions for installing pinned requirements and prepare release by @ml-evs in https://github.com/ppdebreuck/modnet/pull/108

Full Changelog: https://github.com/ppdebreuck/modnet/compare/v0.1.12...v0.1.13

- Jupyter Notebook
Published by ml-evs over 3 years ago

https://github.com/ppdebreuck/modnet - v0.1.12

What's Changed

  • Update MODNet citations by @ml-evs in https://github.com/ppdebreuck/modnet/pull/59
  • add new database for fast featurization by @ppdebreuck in https://github.com/ppdebreuck/modnet/pull/65
  • Rebalance classif by @ppdebreuck in https://github.com/ppdebreuck/modnet/pull/64
  • feature selection updates by @ppdebreuck in https://github.com/ppdebreuck/modnet/pull/58
  • Fit genetic by @gregheymans in https://github.com/ppdebreuck/modnet/pull/50
  • AUC as val metric by @ppdebreuck in https://github.com/ppdebreuck/modnet/pull/68
  • new database + multi-label classification by @ppdebreuck in https://github.com/ppdebreuck/modnet/pull/70
  • num_neurons supposed to be a list with 4 entries (?) by @sgbaird in https://github.com/ppdebreuck/modnet/pull/75
  • Integrity test on saved model by @ppdebreuck in https://github.com/ppdebreuck/modnet/pull/85
  • composition state example (hardness + load), dataset, and citation by @sgbaird in https://github.com/ppdebreuck/modnet/pull/76

New Contributors

  • @dependabot made their first contribution in https://github.com/ppdebreuck/modnet/pull/54
  • @gregheymans made their first contribution in https://github.com/ppdebreuck/modnet/pull/50
  • @sgbaird made their first contribution in https://github.com/ppdebreuck/modnet/pull/75

Full Changelog: https://github.com/ppdebreuck/modnet/compare/v0.1.11...v0.1.12

- Jupyter Notebook
Published by ppdebreuck over 4 years ago

https://github.com/ppdebreuck/modnet - v0.1.11

What's Changed

  • pass classification to kfold split by @ppdebreuck in https://github.com/ppdebreuck/modnet/pull/55
  • A minor fix when running with increasing batch_size scheme by @ml-evs in https://github.com/ppdebreuck/modnet/pull/57

Full Changelog: https://github.com/ppdebreuck/modnet/compare/v0.1.10...v0.1.11

- Jupyter Notebook
Published by ppdebreuck almost 5 years ago

https://github.com/ppdebreuck/modnet - v0.1.10

What's Changed

  • README: add docs badge and move up docs section by @ml-evs in https://github.com/ppdebreuck/modnet/pull/40
  • Adjust ext_data tests to not depickle large MODData by @ml-evs in https://github.com/ppdebreuck/modnet/pull/43
  • Fix benchmark arguments when performing HP optimisation by @ml-evs in https://github.com/ppdebreuck/modnet/pull/41
  • Probabilistic models, fine-grained parallelization for training and feature selection, and smarter data cleaning by @ppdebreuck in https://github.com/ppdebreuck/modnet/pull/44
  • Docs fixes, Add black auto-formatting, linting and pre-commit to CI by @ml-evs in https://github.com/ppdebreuck/modnet/pull/48
  • Post process + classification bug by @ppdebreuck in https://github.com/ppdebreuck/modnet/pull/53

Full Changelog: https://github.com/ppdebreuck/modnet/compare/v0.1.9...v0.1.10

- Jupyter Notebook
Published by ppdebreuck about 5 years ago

https://github.com/ppdebreuck/modnet - v0.1.9

What's Changed

  • v0.1.9 updates from latest preprint: parallel hyperparameter optimisation, nested CV, matbench benchmarking code by @ppdebreuck in https://github.com/ppdebreuck/modnet/pull/23
  • Update README by @ml-evs in https://github.com/ppdebreuck/modnet/pull/29
  • Fix a couple of problems by @ml-evs in https://github.com/ppdebreuck/modnet/pull/30
  • Fix for specifying inner CV fold number by @ml-evs in https://github.com/ppdebreuck/modnet/pull/31
  • fix train-val split when no CV is chosen in fit_preset by @ppdebreuck in https://github.com/ppdebreuck/modnet/pull/33
  • Update tutorials by @ppdebreuck in https://github.com/ppdebreuck/modnet/pull/35
  • Initial docs generation by @ppdebreuck in https://github.com/ppdebreuck/modnet/pull/32

Full Changelog: https://github.com/ppdebreuck/modnet/compare/v0.1.8...v0.1.9

- Jupyter Notebook
Published by ppdebreuck over 5 years ago

https://github.com/ppdebreuck/modnet - v0.1.8

What's Changed

  • Automated validation procedure for hyper parameters by @ppdebreuck in https://github.com/ppdebreuck/modnet/pull/18
  • Refactoring feature selection by @ml-evs in https://github.com/ppdebreuck/modnet/pull/19
  • Refactoring of MODNetModel by @ml-evs in https://github.com/ppdebreuck/modnet/pull/16
  • Tweaking dependencies by @ml-evs in https://github.com/ppdebreuck/modnet/pull/21
  • Classification implementation by @ppdebreuck in https://github.com/ppdebreuck/modnet/pull/20
  • Add ability to load precomputed cross NMI from figshare by @ml-evs in https://github.com/ppdebreuck/modnet/pull/22

Full Changelog: https://github.com/ppdebreuck/modnet/compare/v0.1.7...v0.1.8

- Jupyter Notebook
Published by ml-evs over 5 years ago

https://github.com/ppdebreuck/modnet - v0.1.7

What's Changed

  • Refactored featurization into new module with presets by @ml-evs in https://github.com/ppdebreuck/modnet/pull/15

Full Changelog: https://github.com/ppdebreuck/modnet/compare/0.1.6...v0.1.7

- Jupyter Notebook
Published by ppdebreuck over 5 years ago

https://github.com/ppdebreuck/modnet -

- Jupyter Notebook
Published by ppdebreuck almost 6 years ago

https://github.com/ppdebreuck/modnet -

- Jupyter Notebook
Published by ppdebreuck almost 6 years ago

https://github.com/ppdebreuck/modnet -

- Jupyter Notebook
Published by ppdebreuck about 6 years ago

https://github.com/ppdebreuck/modnet -

- Jupyter Notebook
Published by ppdebreuck about 6 years ago

https://github.com/ppdebreuck/modnet -

- Jupyter Notebook
Published by ppdebreuck about 6 years ago