| Name | neuralGAM JSON |
| Version |
1.0.0
JSON |
| download |
| home_page | https://github.com/inesortega/pyNeuralGAM |
| Summary | Neural network framework based on Generalized Additive Models. |
| upload_time | 2025-03-23 18:07:32 |
| maintainer | None |
| docs_url | None |
| author | Ines Ortega-Fernandez, Marta Sestelo |
| requires_python | >=3.9 |
| license | Mozilla Public License Version 2.0
==================================
This Source Code Form is subject to the terms of the Mozilla Public License, v. 2.0. If a copy of the MPL was not distributed with this file, You can obtain one at https://mozilla.org/MPL/2.0/.
## Software: neuralGAM: Interpretable Neural Network Based on Generalized Additive Models
## Author: Ines Ortega-Fernandez, Marta Sestelo
## Copyright (c) 2025 Ines Ortega-Fernandez
neuralGAM is distributed under the Mozilla Public License 2.0 (MPL 2.0). Under this license:
- You are free to use, modify, and distribute this software.
- Any modifications to the original source code must be made available under MPL 2.0.
- If you distribute this software in executable form, you must also make the source code available under the same license terms.
- The original license, copyright notice, and a link to the MPL must be retained in any copies or substantial portions of the software.
For any additional terms or permissions beyond what is granted under MPL 2.0, you must obtain explicit permission from the author.
This license does not grant you rights to the name, trademark, or branding of neuralGAM. You may not use the neuralGAM name or logo for promotional or commercial purposes without prior consent from the author.
Disclaimer: This software is provided "as-is," without warranty of any kind, express or implied. The author is not liable for any damages arising from the use of this software.
For more information about the Mozilla Public License, please refer to the official documentation: https://mozilla.org/MPL/2.0/.
1. Definitions
--------------
1.1. "Contributor"
means each individual or legal entity that creates, contributes to
the creation of, or owns Covered Software.
1.2. "Contributor Version"
means the combination of the Contributions of others (if any) used
by a Contributor and that particular Contributor's Contribution.
1.3. "Contribution"
means Covered Software of a particular Contributor.
1.4. "Covered Software"
means Source Code Form to which the initial Contributor has attached
the notice in Exhibit A, the Executable Form of such Source Code
Form, and Modifications of such Source Code Form, in each case
including portions thereof.
1.5. "Incompatible With Secondary Licenses"
means
(a) that the initial Contributor has attached the notice described
in Exhibit B to the Covered Software; or
(b) that the Covered Software was made available under the terms of
version 1.1 or earlier of the License, but not also under the
terms of a Secondary License.
1.6. "Executable Form"
means any form of the work other than Source Code Form.
1.7. "Larger Work"
means a work that combines Covered Software with other material, in
a separate file or files, that is not Covered Software.
1.8. "License"
means this document.
1.9. "Licensable"
means having the right to grant, to the maximum extent possible,
whether at the time of the initial grant or subsequently, any and
all of the rights conveyed by this License.
1.10. "Modifications"
means any of the following:
(a) any file in Source Code Form that results from an addition to,
deletion from, or modification of the contents of Covered
Software; or
(b) any new file in Source Code Form that contains any Covered
Software.
1.11. "Patent Claims" of a Contributor
means any patent claim(s), including without limitation, method,
process, and apparatus claims, in any patent Licensable by such
Contributor that would be infringed, but for the grant of the
License, by the making, using, selling, offering for sale, having
made, import, or transfer of either its Contributions or its
Contributor Version.
1.12. "Secondary License"
means either the GNU General Public License, Version 2.0, the GNU
Lesser General Public License, Version 2.1, the GNU Affero General
Public License, Version 3.0, or any later versions of those
licenses.
1.13. "Source Code Form"
means the form of the work preferred for making modifications.
1.14. "You" (or "Your")
means an individual or a legal entity exercising rights under this
License. For legal entities, "You" includes any entity that
controls, is controlled by, or is under common control with You. For
purposes of this definition, "control" means (a) the power, direct
or indirect, to cause the direction or management of such entity,
whether by contract or otherwise, or (b) ownership of more than
fifty percent (50%) of the outstanding shares or beneficial
ownership of such entity.
2. License Grants and Conditions
--------------------------------
2.1. Grants
Each Contributor hereby grants You a world-wide, royalty-free,
non-exclusive license:
(a) under intellectual property rights (other than patent or trademark)
Licensable by such Contributor to use, reproduce, make available,
modify, display, perform, distribute, and otherwise exploit its
Contributions, either on an unmodified basis, with Modifications, or
as part of a Larger Work; and
(b) under Patent Claims of such Contributor to make, use, sell, offer
for sale, have made, import, and otherwise transfer either its
Contributions or its Contributor Version.
2.2. Effective Date
The licenses granted in Section 2.1 with respect to any Contribution
become effective for each Contribution on the date the Contributor first
distributes such Contribution.
2.3. Limitations on Grant Scope
The licenses granted in this Section 2 are the only rights granted under
this License. No additional rights or licenses will be implied from the
distribution or licensing of Covered Software under this License.
Notwithstanding Section 2.1(b) above, no patent license is granted by a
Contributor:
(a) for any code that a Contributor has removed from Covered Software;
or
(b) for infringements caused by: (i) Your and any other third party's
modifications of Covered Software, or (ii) the combination of its
Contributions with other software (except as part of its Contributor
Version); or
(c) under Patent Claims infringed by Covered Software in the absence of
its Contributions.
This License does not grant any rights in the trademarks, service marks,
or logos of any Contributor (except as may be necessary to comply with
the notice requirements in Section 3.4).
2.4. Subsequent Licenses
No Contributor makes additional grants as a result of Your choice to
distribute the Covered Software under a subsequent version of this
License (see Section 10.2) or under the terms of a Secondary License (if
permitted under the terms of Section 3.3).
2.5. Representation
Each Contributor represents that the Contributor believes its
Contributions are its original creation(s) or it has sufficient rights
to grant the rights to its Contributions conveyed by this License.
2.6. Fair Use
This License is not intended to limit any rights You have under
applicable copyright doctrines of fair use, fair dealing, or other
equivalents.
2.7. Conditions
Sections 3.1, 3.2, 3.3, and 3.4 are conditions of the licenses granted
in Section 2.1.
3. Responsibilities
-------------------
3.1. Distribution of Source Form
All distribution of Covered Software in Source Code Form, including any
Modifications that You create or to which You contribute, must be under
the terms of this License. You must inform recipients that the Source
Code Form of the Covered Software is governed by the terms of this
License, and how they can obtain a copy of this License. You may not
attempt to alter or restrict the recipients' rights in the Source Code
Form.
3.2. Distribution of Executable Form
If You distribute Covered Software in Executable Form then:
(a) such Covered Software must also be made available in Source Code
Form, as described in Section 3.1, and You must inform recipients of
the Executable Form how they can obtain a copy of such Source Code
Form by reasonable means in a timely manner, at a charge no more
than the cost of distribution to the recipient; and
(b) You may distribute such Executable Form under the terms of this
License, or sublicense it under different terms, provided that the
license for the Executable Form does not attempt to limit or alter
the recipients' rights in the Source Code Form under this License.
3.3. Distribution of a Larger Work
You may create and distribute a Larger Work under terms of Your choice,
provided that You also comply with the requirements of this License for
the Covered Software. If the Larger Work is a combination of Covered
Software with a work governed by one or more Secondary Licenses, and the
Covered Software is not Incompatible With Secondary Licenses, this
License permits You to additionally distribute such Covered Software
under the terms of such Secondary License(s), so that the recipient of
the Larger Work may, at their option, further distribute the Covered
Software under the terms of either this License or such Secondary
License(s).
3.4. Notices
You may not remove or alter the substance of any license notices
(including copyright notices, patent notices, disclaimers of warranty,
or limitations of liability) contained within the Source Code Form of
the Covered Software, except that You may alter any license notices to
the extent required to remedy known factual inaccuracies.
3.5. Application of Additional Terms
You may choose to offer, and to charge a fee for, warranty, support,
indemnity or liability obligations to one or more recipients of Covered
Software. However, You may do so only on Your own behalf, and not on
behalf of any Contributor. You must make it absolutely clear that any
such warranty, support, indemnity, or liability obligation is offered by
You alone, and You hereby agree to indemnify every Contributor for any
liability incurred by such Contributor as a result of warranty, support,
indemnity or liability terms You offer. You may include additional
disclaimers of warranty and limitations of liability specific to any
jurisdiction.
4. Inability to Comply Due to Statute or Regulation
---------------------------------------------------
If it is impossible for You to comply with any of the terms of this
License with respect to some or all of the Covered Software due to
statute, judicial order, or regulation then You must: (a) comply with
the terms of this License to the maximum extent possible; and (b)
describe the limitations and the code they affect. Such description must
be placed in a text file included with all distributions of the Covered
Software under this License. Except to the extent prohibited by statute
or regulation, such description must be sufficiently detailed for a
recipient of ordinary skill to be able to understand it.
5. Termination
--------------
5.1. The rights granted under this License will terminate automatically
if You fail to comply with any of its terms. However, if You become
compliant, then the rights granted under this License from a particular
Contributor are reinstated (a) provisionally, unless and until such
Contributor explicitly and finally terminates Your grants, and (b) on an
ongoing basis, if such Contributor fails to notify You of the
non-compliance by some reasonable means prior to 60 days after You have
come back into compliance. Moreover, Your grants from a particular
Contributor are reinstated on an ongoing basis if such Contributor
notifies You of the non-compliance by some reasonable means, this is the
first time You have received notice of non-compliance with this License
from such Contributor, and You become compliant prior to 30 days after
Your receipt of the notice.
5.2. If You initiate litigation against any entity by asserting a patent
infringement claim (excluding declaratory judgment actions,
counter-claims, and cross-claims) alleging that a Contributor Version
directly or indirectly infringes any patent, then the rights granted to
You by any and all Contributors for the Covered Software under Section
2.1 of this License shall terminate.
5.3. In the event of termination under Sections 5.1 or 5.2 above, all
end user license agreements (excluding distributors and resellers) which
have been validly granted by You or Your distributors under this License
prior to termination shall survive termination.
6. Disclaimer of Warranty
-------------------------
Covered Software is provided under this License on an "as is" basis, without warranty of any kind, either expressed, implied, or statutory, including, without limitation, warranties that the Covered Software is free of defects, merchantable, fit for a particular purpose or non-infringing. The entire risk as to the quality and performance of the Covered Software is with You. Should any Covered Software prove defective in any respect, You (not any Contributor) assume the cost of any necessary servicing, repair, or correction. This disclaimer of warranty constitutes an essential part of this License. No use of any Covered Software is authorized under this License except under this disclaimer.
7. Limitation of Liability
--------------------------
Under no circumstances and under no legal theory, whether tort (including negligence), contract, or otherwise, shall any Contributor, or anyone who distributes Covered Software as permitted above, be liable to You for any direct, indirect, special, incidental, or consequential damages of any character including, without limitation, damages for lost profits, loss of goodwill, work stoppage, computer failure or malfunction, or any and all other commercial damages or losses, even if such party shall have been informed of the possibility of such damages. This limitation of liability shall not apply to liability for death or personal injury resulting from such party's negligence to the extent applicable law prohibits such limitation. Some jurisdictions do not allow the exclusion or limitation of incidental or consequential damages, so this exclusion and limitation may not apply to You.
8. Litigation
-------------
Any litigation relating to this License may be brought only in the
courts of a jurisdiction where the defendant maintains its principal
place of business and such litigation shall be governed by laws of that
jurisdiction, without reference to its conflict-of-law provisions.
Nothing in this Section shall prevent a party's ability to bring
cross-claims or counter-claims.
9. Miscellaneous
----------------
This License represents the complete agreement concerning the subject
matter hereof. If any provision of this License is held to be
unenforceable, such provision shall be reformed only to the extent
necessary to make it enforceable. Any law or regulation which provides
that the language of a contract shall be construed against the drafter
shall not be used to construe this License against a Contributor.
10. Versions of the License
---------------------------
10.1. New Versions
Mozilla Foundation is the license steward. Except as provided in Section
10.3, no one other than the license steward has the right to modify or
publish new versions of this License. Each version will be given a
distinguishing version number.
10.2. Effect of New Versions
You may distribute the Covered Software under the terms of the version
of the License under which You originally received the Covered Software,
or under the terms of any subsequent version published by the license
steward.
10.3. Modified Versions
If you create software not governed by this License, and you want to
create a new license for such software, you may create and use a
modified version of this License if you rename the license and remove
any references to the name of the license steward (except to note that
such modified license differs from this License).
10.4. Distributing Source Code Form that is Incompatible With Secondary
Licenses
If You choose to distribute Source Code Form that is Incompatible With
Secondary Licenses under the terms of this version of the License, the
notice described in Exhibit B of this License must be attached.
Exhibit A - Source Code Form License Notice
-------------------------------------------
This Source Code Form is subject to the terms of the Mozilla Public
License, v. 2.0. If a copy of the MPL was not distributed with this
file, You can obtain one at http://mozilla.org/MPL/2.0/.
If it is not possible or desirable to put the notice in a particular
file, then You may include the notice in a location (such as a LICENSE
file in a relevant directory) where a recipient would be likely to look
for such a notice.
You may add additional accurate notices of copyright ownership.
Exhibit B - "Incompatible With Secondary Licenses" Notice
---------------------------------------------------------
This Source Code Form is "Incompatible With Secondary Licenses", as
defined by the Mozilla Public License, v. 2.0.
|
| keywords |
deep neural networks
explainable ai
xai
generalized additive models
|
| VCS |
 |
| bugtrack_url |
|
| requirements |
No requirements were recorded.
|
| Travis-CI |
No Travis.
|
| coveralls test coverage |
No coveralls.
|
## neuralGAM: Interpretable Neural Network Based on Generalized Additive Models
Neural Networks are one of the most popular methods nowadays given their high performance on diverse tasks, such as computer vision, anomaly detection, computer-aided disease detection and diagnosis, or natural language processing. However, it is usually unclear how neural networks make decisions, and current methods that try to provide interpretability to neural networks are not robust enough.
We introduce **neuralGAM**, a fully explainable neural network based on **Generalized Additive Models**, which trains a different neural network to estimate the contribution of each feature to the response variable. In contrast to other Neural Additive Models implementations, in **neuralGAM** neural networks are trained independently leveraging the local scoring and backfitting algorithms to ensure that the Generalized Additive Model converges and it is additive. The resultant model is a highly accurate and explainable deep learning model, which can be used for high-risk AI practices where decision-making should be based on accountable and interpretable algorithms.
**neuralGAM** is also available as an R package at the [CRAN](https://cran.r-project.org/package=neuralGAM)
## Installation
To install the neuralGAM package, you can use the following command:
```sh
pip install neuralGAM
```
## Usage
### Linear Regression
To perform linear regression using the neuralGAM package, follow these steps:
1. Import the necessary libraries and the NeuralGAM class:
```python
from neuralGAM.model import NeuralGAM
import pandas as pd
from sklearn.model_selection import train_test_split
from sklearn.metrics import mean_squared_error
```
2. Load your dataset and split it into training and testing sets:
```python
data = pd.read_csv('path/to/your/dataset.csv')
X = data.drop(columns=['target'])
y = data['target']
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
```
3. Initialize the NeuralGAM model. You might need to adjust the `num_units` parameter depending on your data complexity and availability. Each number in the list defines the number of hidden units in each layer of the Deep Neural Network. For simple problems we recommend a single-layer neural network with 1024 units.
```python
ngam = NeuralGAM(family="gaussian", num_units=[1024], learning_rate=0.00053)
```
4. Fit the model to the training data:
```python
muhat, fs_train_estimated, eta = ngam.fit(X_train=X_train, y_train=y_train, max_iter_ls=10, bf_threshold=1e-5, ls_threshold=0.01, max_iter_backfitting=10, parallel=True)
```
5. Make predictions on the test data and compute the mean squared error:
```python
y_pred = ngam.predict(X_test, type="response")
pred_err = mean_squared_error(y_test, y_pred)
print(f"MSE in the test set = {pred_err}")
```
6. Plot the partial dependencies:
```python
from neuralGAM.plot import plot_partial_dependencies
import matplotlib.pyplot as plt
plt.style.use('seaborn-v0_8')
plot_partial_dependencies(x=X_train, fs=fs_train_estimated, title="Estimated Training Partial Effects")
fs_test_est = ngam.predict(X_test, type="terms")
plot_partial_dependencies(x=X_test, fs=fs_test_est, title="Estimated Test Partial Effects")
```
### Logistic Regression
To perform logistic regression using the neuralGAM package, follow these steps:
1. Import the necessary libraries and the NeuralGAM class:
```python
from neuralGAM.model import NeuralGAM
import pandas as pd
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score
```
2. Load your dataset and split it into training and testing sets:
```python
data = pd.read_csv('path/to/your/dataset.csv')
X = data.drop(columns=['target'])
y = data['target']
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
```
3. Initialize the NeuralGAM model. You might need to adjust the `num_units` parameter depending on your data complexity and availability. Each number in the list defines the number of hidden units in each layer of the Deep Neural Network. For simple problems we recommend a single-layer neural network with 1024 units. If you want to force a linear fit for a specific covariate, you can do so using the `linear_terms` parameter:
```python
ngam = NeuralGAM(family="binomial", num_units=[1024], learning_rate=0.00053, linear_terms=[1])
```
4. Fit the model to the training data:
```python
muhat, fs_train_estimated, eta = ngam.fit(X_train=X_train, y_train=y_train, max_iter_ls=10, bf_threshold=1e-5, ls_threshold=0.01, max_iter_backfitting=10, parallel=True)
```
5. Make predictions on the test data and compute the accuracy:
```python
y_pred = ngam.predict(X_test, type="response") # get predicted probabilities
y_pred_class = (y_pred > 0.5).astype(int)
accuracy = accuracy_score(y_test, y_pred_class) # assuming y_test is in the discrete set {0,1}
print(f"Accuracy in the test set = {accuracy}")
```
6. Plot the partial dependencies:
```python
from neuralGAM.plot import plot_partial_dependencies
import matplotlib.pyplot as plt
plt.style.use('seaborn-v0_8')
plot_partial_dependencies(x=X_train, fs=fs_train_estimated, title="Estimated Training Partial Effects")
fs_test_est = ngam.predict(X_test, type="terms")
plot_partial_dependencies(x=X_test, fs=fs_test_est, title="Estimated Test Partial Effects")
```
## Examples
You can find detailed examples for both linear and logistic regression in the examples folder. These examples are provided as Jupyter notebooks:
- Linear Regression Example
- Logistic Regression Example
## Citation
If you use neuralGAM in your research, please cite the following paper:
> Ortega-Fernandez, I., Sestelo, M. & Villanueva, N.M. Explainable generalized additive neural networks with independent neural network training. Stat Comput 34, 6 (2024). https://doi.org/10.1007/s11222-023-10320-5
```bibtex
@article{ortega2024explainable,
title={Explainable generalized additive neural networks with independent neural network training},
author={Ortega-Fernandez, Ines and Sestelo, Marta and Villanueva, Nora M},
journal={Statistics and Computing},
volume={34},
number={1},
pages={6},
year={2024},
publisher={Springer}
}
```
Raw data
{
"_id": null,
"home_page": "https://github.com/inesortega/pyNeuralGAM",
"name": "neuralGAM",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.9",
"maintainer_email": "Ines Ortega-Fernandez <iortega@gradiant.org>",
"keywords": "deep neural networks, explainable ai, xAI, Generalized Additive Models",
"author": "Ines Ortega-Fernandez, Marta Sestelo",
"author_email": "Ines Ortega-Fernandez <iortega@gradiant.org>, Marta Sestelo <sestelo@uvigo.gal>",
"download_url": "https://files.pythonhosted.org/packages/d8/98/74d5ca61200a9b94aeb32951d5f3dfb3bd94cd8407ab9a67c10bec9a2114/neuralgam-1.0.0.tar.gz",
"platform": null,
"description": "## neuralGAM: Interpretable Neural Network Based on Generalized Additive Models\n\nNeural Networks are one of the most popular methods nowadays given their high performance on diverse tasks, such as computer vision, anomaly detection, computer-aided disease detection and diagnosis, or natural language processing. However, it is usually unclear how neural networks make decisions, and current methods that try to provide interpretability to neural networks are not robust enough.\n\nWe introduce **neuralGAM**, a fully explainable neural network based on **Generalized Additive Models**, which trains a different neural network to estimate the contribution of each feature to the response variable. In contrast to other Neural Additive Models implementations, in **neuralGAM** neural networks are trained independently leveraging the local scoring and backfitting algorithms to ensure that the Generalized Additive Model converges and it is additive. The resultant model is a highly accurate and explainable deep learning model, which can be used for high-risk AI practices where decision-making should be based on accountable and interpretable algorithms.\n\n**neuralGAM** is also available as an R package at the [CRAN](https://cran.r-project.org/package=neuralGAM)\n\n## Installation\n\nTo install the neuralGAM package, you can use the following command:\n\n```sh\npip install neuralGAM\n```\n\n## Usage\n\n### Linear Regression\n\nTo perform linear regression using the neuralGAM package, follow these steps:\n\n1. Import the necessary libraries and the NeuralGAM class:\n\n ```python\n from neuralGAM.model import NeuralGAM\n import pandas as pd\n from sklearn.model_selection import train_test_split\n from sklearn.metrics import mean_squared_error\n ```\n\n2. Load your dataset and split it into training and testing sets:\n\n ```python\n data = pd.read_csv('path/to/your/dataset.csv')\n X = data.drop(columns=['target'])\n y = data['target']\n X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)\n ```\n\n3. Initialize the NeuralGAM model. You might need to adjust the `num_units` parameter depending on your data complexity and availability. Each number in the list defines the number of hidden units in each layer of the Deep Neural Network. For simple problems we recommend a single-layer neural network with 1024 units.\n\n ```python\n ngam = NeuralGAM(family=\"gaussian\", num_units=[1024], learning_rate=0.00053)\n ```\n\n4. Fit the model to the training data:\n\n ```python\n muhat, fs_train_estimated, eta = ngam.fit(X_train=X_train, y_train=y_train, max_iter_ls=10, bf_threshold=1e-5, ls_threshold=0.01, max_iter_backfitting=10, parallel=True)\n ```\n\n5. Make predictions on the test data and compute the mean squared error:\n\n ```python\n y_pred = ngam.predict(X_test, type=\"response\")\n pred_err = mean_squared_error(y_test, y_pred)\n print(f\"MSE in the test set = {pred_err}\")\n ```\n\n6. Plot the partial dependencies:\n\n ```python\n from neuralGAM.plot import plot_partial_dependencies\n import matplotlib.pyplot as plt\n\n plt.style.use('seaborn-v0_8')\n plot_partial_dependencies(x=X_train, fs=fs_train_estimated, title=\"Estimated Training Partial Effects\")\n fs_test_est = ngam.predict(X_test, type=\"terms\")\n plot_partial_dependencies(x=X_test, fs=fs_test_est, title=\"Estimated Test Partial Effects\")\n ```\n\n### Logistic Regression\n\nTo perform logistic regression using the neuralGAM package, follow these steps:\n\n1. Import the necessary libraries and the NeuralGAM class:\n\n ```python\n from neuralGAM.model import NeuralGAM\n import pandas as pd\n from sklearn.model_selection import train_test_split\n from sklearn.metrics import accuracy_score\n ```\n\n2. Load your dataset and split it into training and testing sets:\n\n ```python\n data = pd.read_csv('path/to/your/dataset.csv')\n X = data.drop(columns=['target'])\n y = data['target']\n X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)\n ```\n\n3. Initialize the NeuralGAM model. You might need to adjust the `num_units` parameter depending on your data complexity and availability. Each number in the list defines the number of hidden units in each layer of the Deep Neural Network. For simple problems we recommend a single-layer neural network with 1024 units. If you want to force a linear fit for a specific covariate, you can do so using the `linear_terms` parameter:\n\n ```python\n ngam = NeuralGAM(family=\"binomial\", num_units=[1024], learning_rate=0.00053, linear_terms=[1])\n ```\n\n4. Fit the model to the training data:\n\n ```python\n muhat, fs_train_estimated, eta = ngam.fit(X_train=X_train, y_train=y_train, max_iter_ls=10, bf_threshold=1e-5, ls_threshold=0.01, max_iter_backfitting=10, parallel=True)\n ```\n\n5. Make predictions on the test data and compute the accuracy:\n\n ```python\n y_pred = ngam.predict(X_test, type=\"response\") # get predicted probabilities\n y_pred_class = (y_pred > 0.5).astype(int) \n accuracy = accuracy_score(y_test, y_pred_class) # assuming y_test is in the discrete set {0,1}\n print(f\"Accuracy in the test set = {accuracy}\")\n ```\n\n6. Plot the partial dependencies:\n\n ```python\n from neuralGAM.plot import plot_partial_dependencies\n import matplotlib.pyplot as plt\n\n plt.style.use('seaborn-v0_8')\n plot_partial_dependencies(x=X_train, fs=fs_train_estimated, title=\"Estimated Training Partial Effects\")\n fs_test_est = ngam.predict(X_test, type=\"terms\")\n plot_partial_dependencies(x=X_test, fs=fs_test_est, title=\"Estimated Test Partial Effects\")\n ```\n\n## Examples\n\nYou can find detailed examples for both linear and logistic regression in the examples folder. These examples are provided as Jupyter notebooks:\n\n- Linear Regression Example\n- Logistic Regression Example\n\n## Citation\n\nIf you use neuralGAM in your research, please cite the following paper:\n\n> Ortega-Fernandez, I., Sestelo, M. & Villanueva, N.M. Explainable generalized additive neural networks with independent neural network training. Stat Comput 34, 6 (2024). https://doi.org/10.1007/s11222-023-10320-5\n\n```bibtex\n@article{ortega2024explainable,\n title={Explainable generalized additive neural networks with independent neural network training},\n author={Ortega-Fernandez, Ines and Sestelo, Marta and Villanueva, Nora M},\n journal={Statistics and Computing},\n volume={34},\n number={1},\n pages={6},\n year={2024},\n publisher={Springer}\n}\n```\n",
"bugtrack_url": null,
"license": "Mozilla Public License Version 2.0\n ==================================\n \n This Source Code Form is subject to the terms of the Mozilla Public License, v. 2.0. If a copy of the MPL was not distributed with this file, You can obtain one at https://mozilla.org/MPL/2.0/.\n \n ## Software: neuralGAM: Interpretable Neural Network Based on Generalized Additive Models\n ## Author: Ines Ortega-Fernandez, Marta Sestelo\n ## Copyright (c) 2025 Ines Ortega-Fernandez\n \n neuralGAM is distributed under the Mozilla Public License 2.0 (MPL 2.0). Under this license:\n - You are free to use, modify, and distribute this software.\n - Any modifications to the original source code must be made available under MPL 2.0.\n - If you distribute this software in executable form, you must also make the source code available under the same license terms.\n - The original license, copyright notice, and a link to the MPL must be retained in any copies or substantial portions of the software.\n \n For any additional terms or permissions beyond what is granted under MPL 2.0, you must obtain explicit permission from the author.\n \n This license does not grant you rights to the name, trademark, or branding of neuralGAM. You may not use the neuralGAM name or logo for promotional or commercial purposes without prior consent from the author.\n \n Disclaimer: This software is provided \"as-is,\" without warranty of any kind, express or implied. The author is not liable for any damages arising from the use of this software.\n \n For more information about the Mozilla Public License, please refer to the official documentation: https://mozilla.org/MPL/2.0/.\n \n 1. Definitions\n --------------\n \n 1.1. \"Contributor\"\n means each individual or legal entity that creates, contributes to\n the creation of, or owns Covered Software.\n \n 1.2. \"Contributor Version\"\n means the combination of the Contributions of others (if any) used\n by a Contributor and that particular Contributor's Contribution.\n \n 1.3. \"Contribution\"\n means Covered Software of a particular Contributor.\n \n 1.4. \"Covered Software\"\n means Source Code Form to which the initial Contributor has attached\n the notice in Exhibit A, the Executable Form of such Source Code\n Form, and Modifications of such Source Code Form, in each case\n including portions thereof.\n \n 1.5. \"Incompatible With Secondary Licenses\"\n means\n \n (a) that the initial Contributor has attached the notice described\n in Exhibit B to the Covered Software; or\n \n (b) that the Covered Software was made available under the terms of\n version 1.1 or earlier of the License, but not also under the\n terms of a Secondary License.\n \n 1.6. \"Executable Form\"\n means any form of the work other than Source Code Form.\n \n 1.7. \"Larger Work\"\n means a work that combines Covered Software with other material, in\n a separate file or files, that is not Covered Software.\n \n 1.8. \"License\"\n means this document.\n \n 1.9. \"Licensable\"\n means having the right to grant, to the maximum extent possible,\n whether at the time of the initial grant or subsequently, any and\n all of the rights conveyed by this License.\n \n 1.10. \"Modifications\"\n means any of the following:\n \n (a) any file in Source Code Form that results from an addition to,\n deletion from, or modification of the contents of Covered\n Software; or\n \n (b) any new file in Source Code Form that contains any Covered\n Software.\n \n 1.11. \"Patent Claims\" of a Contributor\n means any patent claim(s), including without limitation, method,\n process, and apparatus claims, in any patent Licensable by such\n Contributor that would be infringed, but for the grant of the\n License, by the making, using, selling, offering for sale, having\n made, import, or transfer of either its Contributions or its\n Contributor Version.\n \n 1.12. \"Secondary License\"\n means either the GNU General Public License, Version 2.0, the GNU\n Lesser General Public License, Version 2.1, the GNU Affero General\n Public License, Version 3.0, or any later versions of those\n licenses.\n \n 1.13. \"Source Code Form\"\n means the form of the work preferred for making modifications.\n \n 1.14. \"You\" (or \"Your\")\n means an individual or a legal entity exercising rights under this\n License. For legal entities, \"You\" includes any entity that\n controls, is controlled by, or is under common control with You. For\n purposes of this definition, \"control\" means (a) the power, direct\n or indirect, to cause the direction or management of such entity,\n whether by contract or otherwise, or (b) ownership of more than\n fifty percent (50%) of the outstanding shares or beneficial\n ownership of such entity.\n \n 2. License Grants and Conditions\n --------------------------------\n \n 2.1. Grants\n \n Each Contributor hereby grants You a world-wide, royalty-free,\n non-exclusive license:\n \n (a) under intellectual property rights (other than patent or trademark)\n Licensable by such Contributor to use, reproduce, make available,\n modify, display, perform, distribute, and otherwise exploit its\n Contributions, either on an unmodified basis, with Modifications, or\n as part of a Larger Work; and\n \n (b) under Patent Claims of such Contributor to make, use, sell, offer\n for sale, have made, import, and otherwise transfer either its\n Contributions or its Contributor Version.\n \n 2.2. Effective Date\n \n The licenses granted in Section 2.1 with respect to any Contribution\n become effective for each Contribution on the date the Contributor first\n distributes such Contribution.\n \n 2.3. Limitations on Grant Scope\n \n The licenses granted in this Section 2 are the only rights granted under\n this License. No additional rights or licenses will be implied from the\n distribution or licensing of Covered Software under this License.\n Notwithstanding Section 2.1(b) above, no patent license is granted by a\n Contributor:\n \n (a) for any code that a Contributor has removed from Covered Software;\n or\n \n (b) for infringements caused by: (i) Your and any other third party's\n modifications of Covered Software, or (ii) the combination of its\n Contributions with other software (except as part of its Contributor\n Version); or\n \n (c) under Patent Claims infringed by Covered Software in the absence of\n its Contributions.\n \n This License does not grant any rights in the trademarks, service marks,\n or logos of any Contributor (except as may be necessary to comply with\n the notice requirements in Section 3.4).\n \n 2.4. Subsequent Licenses\n \n No Contributor makes additional grants as a result of Your choice to\n distribute the Covered Software under a subsequent version of this\n License (see Section 10.2) or under the terms of a Secondary License (if\n permitted under the terms of Section 3.3).\n \n 2.5. Representation\n \n Each Contributor represents that the Contributor believes its\n Contributions are its original creation(s) or it has sufficient rights\n to grant the rights to its Contributions conveyed by this License.\n \n 2.6. Fair Use\n \n This License is not intended to limit any rights You have under\n applicable copyright doctrines of fair use, fair dealing, or other\n equivalents.\n \n 2.7. Conditions\n \n Sections 3.1, 3.2, 3.3, and 3.4 are conditions of the licenses granted\n in Section 2.1.\n \n 3. Responsibilities\n -------------------\n \n 3.1. Distribution of Source Form\n \n All distribution of Covered Software in Source Code Form, including any\n Modifications that You create or to which You contribute, must be under\n the terms of this License. You must inform recipients that the Source\n Code Form of the Covered Software is governed by the terms of this\n License, and how they can obtain a copy of this License. You may not\n attempt to alter or restrict the recipients' rights in the Source Code\n Form.\n \n 3.2. Distribution of Executable Form\n \n If You distribute Covered Software in Executable Form then:\n \n (a) such Covered Software must also be made available in Source Code\n Form, as described in Section 3.1, and You must inform recipients of\n the Executable Form how they can obtain a copy of such Source Code\n Form by reasonable means in a timely manner, at a charge no more\n than the cost of distribution to the recipient; and\n \n (b) You may distribute such Executable Form under the terms of this\n License, or sublicense it under different terms, provided that the\n license for the Executable Form does not attempt to limit or alter\n the recipients' rights in the Source Code Form under this License.\n \n 3.3. Distribution of a Larger Work\n \n You may create and distribute a Larger Work under terms of Your choice,\n provided that You also comply with the requirements of this License for\n the Covered Software. If the Larger Work is a combination of Covered\n Software with a work governed by one or more Secondary Licenses, and the\n Covered Software is not Incompatible With Secondary Licenses, this\n License permits You to additionally distribute such Covered Software\n under the terms of such Secondary License(s), so that the recipient of\n the Larger Work may, at their option, further distribute the Covered\n Software under the terms of either this License or such Secondary\n License(s).\n \n 3.4. Notices\n \n You may not remove or alter the substance of any license notices\n (including copyright notices, patent notices, disclaimers of warranty,\n or limitations of liability) contained within the Source Code Form of\n the Covered Software, except that You may alter any license notices to\n the extent required to remedy known factual inaccuracies.\n \n 3.5. Application of Additional Terms\n \n You may choose to offer, and to charge a fee for, warranty, support,\n indemnity or liability obligations to one or more recipients of Covered\n Software. However, You may do so only on Your own behalf, and not on\n behalf of any Contributor. You must make it absolutely clear that any\n such warranty, support, indemnity, or liability obligation is offered by\n You alone, and You hereby agree to indemnify every Contributor for any\n liability incurred by such Contributor as a result of warranty, support,\n indemnity or liability terms You offer. You may include additional\n disclaimers of warranty and limitations of liability specific to any\n jurisdiction.\n \n 4. Inability to Comply Due to Statute or Regulation\n ---------------------------------------------------\n \n If it is impossible for You to comply with any of the terms of this\n License with respect to some or all of the Covered Software due to\n statute, judicial order, or regulation then You must: (a) comply with\n the terms of this License to the maximum extent possible; and (b)\n describe the limitations and the code they affect. Such description must\n be placed in a text file included with all distributions of the Covered\n Software under this License. Except to the extent prohibited by statute\n or regulation, such description must be sufficiently detailed for a\n recipient of ordinary skill to be able to understand it.\n \n 5. Termination\n --------------\n \n 5.1. The rights granted under this License will terminate automatically\n if You fail to comply with any of its terms. However, if You become\n compliant, then the rights granted under this License from a particular\n Contributor are reinstated (a) provisionally, unless and until such\n Contributor explicitly and finally terminates Your grants, and (b) on an\n ongoing basis, if such Contributor fails to notify You of the\n non-compliance by some reasonable means prior to 60 days after You have\n come back into compliance. Moreover, Your grants from a particular\n Contributor are reinstated on an ongoing basis if such Contributor\n notifies You of the non-compliance by some reasonable means, this is the\n first time You have received notice of non-compliance with this License\n from such Contributor, and You become compliant prior to 30 days after\n Your receipt of the notice.\n \n 5.2. If You initiate litigation against any entity by asserting a patent\n infringement claim (excluding declaratory judgment actions,\n counter-claims, and cross-claims) alleging that a Contributor Version\n directly or indirectly infringes any patent, then the rights granted to\n You by any and all Contributors for the Covered Software under Section\n 2.1 of this License shall terminate.\n \n 5.3. In the event of termination under Sections 5.1 or 5.2 above, all\n end user license agreements (excluding distributors and resellers) which\n have been validly granted by You or Your distributors under this License\n prior to termination shall survive termination.\n \n 6. Disclaimer of Warranty \n -------------------------\n Covered Software is provided under this License on an \"as is\" basis, without warranty of any kind, either expressed, implied, or statutory, including, without limitation, warranties that the Covered Software is free of defects, merchantable, fit for a particular purpose or non-infringing. The entire risk as to the quality and performance of the Covered Software is with You. Should any Covered Software prove defective in any respect, You (not any Contributor) assume the cost of any necessary servicing, repair, or correction. This disclaimer of warranty constitutes an essential part of this License. No use of any Covered Software is authorized under this License except under this disclaimer.\n \n 7. Limitation of Liability \n -------------------------- \n Under no circumstances and under no legal theory, whether tort (including negligence), contract, or otherwise, shall any Contributor, or anyone who distributes Covered Software as permitted above, be liable to You for any direct, indirect, special, incidental, or consequential damages of any character including, without limitation, damages for lost profits, loss of goodwill, work stoppage, computer failure or malfunction, or any and all other commercial damages or losses, even if such party shall have been informed of the possibility of such damages. This limitation of liability shall not apply to liability for death or personal injury resulting from such party's negligence to the extent applicable law prohibits such limitation. Some jurisdictions do not allow the exclusion or limitation of incidental or consequential damages, so this exclusion and limitation may not apply to You.\n \n 8. Litigation\n -------------\n \n Any litigation relating to this License may be brought only in the\n courts of a jurisdiction where the defendant maintains its principal\n place of business and such litigation shall be governed by laws of that\n jurisdiction, without reference to its conflict-of-law provisions.\n Nothing in this Section shall prevent a party's ability to bring\n cross-claims or counter-claims.\n \n 9. Miscellaneous\n ----------------\n \n This License represents the complete agreement concerning the subject\n matter hereof. If any provision of this License is held to be\n unenforceable, such provision shall be reformed only to the extent\n necessary to make it enforceable. Any law or regulation which provides\n that the language of a contract shall be construed against the drafter\n shall not be used to construe this License against a Contributor.\n \n 10. Versions of the License\n ---------------------------\n \n 10.1. New Versions\n \n Mozilla Foundation is the license steward. Except as provided in Section\n 10.3, no one other than the license steward has the right to modify or\n publish new versions of this License. Each version will be given a\n distinguishing version number.\n \n 10.2. Effect of New Versions\n \n You may distribute the Covered Software under the terms of the version\n of the License under which You originally received the Covered Software,\n or under the terms of any subsequent version published by the license\n steward.\n \n 10.3. Modified Versions\n \n If you create software not governed by this License, and you want to\n create a new license for such software, you may create and use a\n modified version of this License if you rename the license and remove\n any references to the name of the license steward (except to note that\n such modified license differs from this License).\n \n 10.4. Distributing Source Code Form that is Incompatible With Secondary\n Licenses\n \n If You choose to distribute Source Code Form that is Incompatible With\n Secondary Licenses under the terms of this version of the License, the\n notice described in Exhibit B of this License must be attached.\n \n Exhibit A - Source Code Form License Notice\n -------------------------------------------\n \n This Source Code Form is subject to the terms of the Mozilla Public\n License, v. 2.0. If a copy of the MPL was not distributed with this\n file, You can obtain one at http://mozilla.org/MPL/2.0/.\n \n If it is not possible or desirable to put the notice in a particular\n file, then You may include the notice in a location (such as a LICENSE\n file in a relevant directory) where a recipient would be likely to look\n for such a notice.\n \n You may add additional accurate notices of copyright ownership.\n \n Exhibit B - \"Incompatible With Secondary Licenses\" Notice\n ---------------------------------------------------------\n \n This Source Code Form is \"Incompatible With Secondary Licenses\", as\n defined by the Mozilla Public License, v. 2.0.\n ",
"summary": "Neural network framework based on Generalized Additive Models.",
"version": "1.0.0",
"project_urls": {
"Homepage": "https://inesortega.github.io/software/2023-neuralGAM",
"Issues": "https://github.com/inesortega/pyNeuralGAM/issues",
"Repository": "https://github.com/inesortega/pyneuralGAM"
},
"split_keywords": [
"deep neural networks",
" explainable ai",
" xai",
" generalized additive models"
],
"urls": [
{
"comment_text": null,
"digests": {
"blake2b_256": "a978137fb82978723829aa109a33192da2e18ee44fcad22b6056705c5857c96c",
"md5": "747e3d16bfea353259e08b6f5b34ef04",
"sha256": "0d583e5bba2dde40d588ac01f728afb7e0f5ea8045fa396f1330b7dd1afa0822"
},
"downloads": -1,
"filename": "neuralgam-1.0.0-py3-none-any.whl",
"has_sig": false,
"md5_digest": "747e3d16bfea353259e08b6f5b34ef04",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.9",
"size": 32370,
"upload_time": "2025-03-23T18:07:30",
"upload_time_iso_8601": "2025-03-23T18:07:30.258450Z",
"url": "https://files.pythonhosted.org/packages/a9/78/137fb82978723829aa109a33192da2e18ee44fcad22b6056705c5857c96c/neuralgam-1.0.0-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": null,
"digests": {
"blake2b_256": "d89874d5ca61200a9b94aeb32951d5f3dfb3bd94cd8407ab9a67c10bec9a2114",
"md5": "10b9ceac304ef29bb5d3c793d22f35da",
"sha256": "b07e93acea9bcce72d3837f8a4dba999f2a063ffef05da867dbbc8dbddb141dc"
},
"downloads": -1,
"filename": "neuralgam-1.0.0.tar.gz",
"has_sig": false,
"md5_digest": "10b9ceac304ef29bb5d3c793d22f35da",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.9",
"size": 29290,
"upload_time": "2025-03-23T18:07:32",
"upload_time_iso_8601": "2025-03-23T18:07:32.095683Z",
"url": "https://files.pythonhosted.org/packages/d8/98/74d5ca61200a9b94aeb32951d5f3dfb3bd94cd8407ab9a67c10bec9a2114/neuralgam-1.0.0.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2025-03-23 18:07:32",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "inesortega",
"github_project": "pyNeuralGAM",
"travis_ci": false,
"coveralls": false,
"github_actions": false,
"lcname": "neuralgam"
}
