Interpretability module: sparse linear models via LASSO

docs/api.rst

@@ -33,10 +33,18 @@ Here we provide PyTorch- and Lightning-compatible classes that simplify and impr
 
 .. rubric:: Utils
 
-Here one can find a set of miscellaneous tools and helper functions for a smoother workflow.
+Miscellaneous tools and helper functions:
 
 .. toctree::
    :maxdepth: 1
    :caption: API:
 
    api_utils
+
+Tools for inspecting & interpreting the models.
+
+.. toctree::
+   :maxdepth: 1
+   :caption: API:
+
+   api_explain

docs/api_cvs.rst

@@ -71,6 +71,8 @@ General framework which allows to optimize a single model with different loss fu
 
 Framework for the numerical determination of the committor function based on its variational principle.
 
+.. rubric:: Committor 
+
 .. currentmodule:: mlcolvar.cvs
 
 .. autosummary::

docs/api_explain.rst

@@ -0,0 +1,30 @@
+Explain
+-------
+
+.. rubric:: Sensitivity analysis
+
+Perform sensitivity analysis to identify feature relevances
+
+.. currentmodule:: mlcolvar.explain.sensitivity
+
+.. autosummary::
+   :toctree: autosummary
+   :template: custom-class-template.rst
+
+   sensitivity_analysis 
+   plot_sensitivity
+
+.. rubric:: Sparse linear model
+
+Use sparse models to approximate classification/regression tasks
+
+.. currentmodule:: mlcolvar.explain.lasso
+
+.. autosummary::
+   :toctree: autosummary
+   :template: custom-class-template.rst
+
+   lasso_classification
+   lasso_regression
+   plot_lasso_classification
+   plot_lasso_regression

docs/api_utils.rst

@@ -28,6 +28,18 @@ Create a dataset of pairs of time-lagged configurations.
 
 .. rubric:: FES
 
+.. rubric:: Trainer
+
+Functions used in conjunction with the lightning Trainer (e.g. logging, metrics...).
+
+.. currentmodule:: mlcolvar.utils.trainer
+
+.. autosummary::
+   :toctree: autosummary
+   :template: custom-class-template.rst
+
+   MetricsCallback
+
 Compute (and plot) the free energy surface along the CVs.
 
 .. currentmodule:: mlcolvar.utils.fes
@@ -38,14 +50,13 @@ Compute (and plot) the free energy surface along the CVs.
 
    compute_fes
 
-.. rubric:: Trainer
+Plotting utils
 
-Functions used in conjunction with the lightning Trainer (e.g. logging, metrics...).
-
-.. currentmodule:: mlcolvar.utils.trainer
+.. currentmodule:: mlcolvar.utils.plot
 
 .. autosummary::
    :toctree: autosummary
    :template: custom-class-template.rst
 
-   MetricsCallback
+   plot_metrics
+   plot_features_distribution

docs/autosummary/mlcolvar.core.transform.descriptors.CoordinationNumbers.rst

@@ -0,0 +1,40 @@
+mlcolvar.core.transform.descriptors.CoordinationNumbers
+=======================================================
+
+.. currentmodule:: mlcolvar.core.transform.descriptors
+
+.. autoclass:: CoordinationNumbers
+   :members:                                   
+   :show-inheritance:                           
+   :inherited-members: Module,LightningModule                       
+
+
+   .. automethod:: __init__
+
+
+   .. rubric:: Methods
+
+   .. autosummary::
+
+      ~CoordinationNumbers.__init__
+      ~CoordinationNumbers.compute_coordination_number
+      ~CoordinationNumbers.forward
+
+
+
+
+..
+
+
+   .. rubric:: Attributes
+
+   .. autosummary::
+
+      ~CoordinationNumbers.T_destination
+      ~CoordinationNumbers.call_super_init
+      ~CoordinationNumbers.dump_patches
+      ~CoordinationNumbers.training
+
+
+
+

docs/autosummary/mlcolvar.core.transform.descriptors.EigsAdjMat.rst

@@ -0,0 +1,41 @@
+mlcolvar.core.transform.descriptors.EigsAdjMat
+==============================================
+
+.. currentmodule:: mlcolvar.core.transform.descriptors
+
+.. autoclass:: EigsAdjMat
+   :members:                                   
+   :show-inheritance:                           
+   :inherited-members: Module,LightningModule                       
+
+
+   .. automethod:: __init__
+
+
+   .. rubric:: Methods
+
+   .. autosummary::
+
+      ~EigsAdjMat.__init__
+      ~EigsAdjMat.compute_adjacency_matrix
+      ~EigsAdjMat.forward
+      ~EigsAdjMat.get_eigenvalues
+
+
+
+
+..
+
+
+   .. rubric:: Attributes
+
+   .. autosummary::
+
+      ~EigsAdjMat.T_destination
+      ~EigsAdjMat.call_super_init
+      ~EigsAdjMat.dump_patches
+      ~EigsAdjMat.training
+
+
+
+

docs/autosummary/mlcolvar.core.transform.descriptors.MultipleDescriptors.rst

@@ -0,0 +1,39 @@
+mlcolvar.core.transform.descriptors.MultipleDescriptors
+=======================================================
+
+.. currentmodule:: mlcolvar.core.transform.descriptors
+
+.. autoclass:: MultipleDescriptors
+   :members:                                   
+   :show-inheritance:                           
+   :inherited-members: Module,LightningModule                       
+
+
+   .. automethod:: __init__
+
+
+   .. rubric:: Methods
+
+   .. autosummary::
+
+      ~MultipleDescriptors.__init__
+      ~MultipleDescriptors.forward
+
+
+
+
+..
+
+
+   .. rubric:: Attributes
+
+   .. autosummary::
+
+      ~MultipleDescriptors.T_destination
+      ~MultipleDescriptors.call_super_init
+      ~MultipleDescriptors.dump_patches
+      ~MultipleDescriptors.training
+
+
+
+

docs/autosummary/mlcolvar.core.transform.descriptors.PairwiseDistances.rst

@@ -0,0 +1,40 @@
+mlcolvar.core.transform.descriptors.PairwiseDistances
+=====================================================
+
+.. currentmodule:: mlcolvar.core.transform.descriptors
+
+.. autoclass:: PairwiseDistances
+   :members:                                   
+   :show-inheritance:                           
+   :inherited-members: Module,LightningModule                       
+
+
+   .. automethod:: __init__
+
+
+   .. rubric:: Methods
+
+   .. autosummary::
+
+      ~PairwiseDistances.__init__
+      ~PairwiseDistances.compute_pairwise_distances
+      ~PairwiseDistances.forward
+
+
+
+
+..
+
+
+   .. rubric:: Attributes
+
+   .. autosummary::
+
+      ~PairwiseDistances.T_destination
+      ~PairwiseDistances.call_super_init
+      ~PairwiseDistances.dump_patches
+      ~PairwiseDistances.training
+
+
+
+

docs/autosummary/mlcolvar.core.transform.descriptors.TorsionalAngle.rst

@@ -0,0 +1,41 @@
+mlcolvar.core.transform.descriptors.TorsionalAngle
+==================================================
+
+.. currentmodule:: mlcolvar.core.transform.descriptors
+
+.. autoclass:: TorsionalAngle
+   :members:                                   
+   :show-inheritance:                           
+   :inherited-members: Module,LightningModule                       
+
+
+   .. automethod:: __init__
+
+
+   .. rubric:: Methods
+
+   .. autosummary::
+
+      ~TorsionalAngle.__init__
+      ~TorsionalAngle.compute_torsional_angle
+      ~TorsionalAngle.forward
+
+
+
+
+..
+
+
+   .. rubric:: Attributes
+
+   .. autosummary::
+
+      ~TorsionalAngle.MODES
+      ~TorsionalAngle.T_destination
+      ~TorsionalAngle.call_super_init
+      ~TorsionalAngle.dump_patches
+      ~TorsionalAngle.training
+
+
+
+

docs/autosummary/mlcolvar.core.transform.tools.ContinuousHistogram.rst

@@ -0,0 +1,40 @@
+mlcolvar.core.transform.tools.ContinuousHistogram
+=================================================
+
+.. currentmodule:: mlcolvar.core.transform.tools
+
+.. autoclass:: ContinuousHistogram
+   :members:                                   
+   :show-inheritance:                           
+   :inherited-members: Module,LightningModule                       
+
+
+   .. automethod:: __init__
+
+
+   .. rubric:: Methods
+
+   .. autosummary::
+
+      ~ContinuousHistogram.__init__
+      ~ContinuousHistogram.compute_hist
+      ~ContinuousHistogram.forward
+
+
+
+
+..
+
+
+   .. rubric:: Attributes
+
+   .. autosummary::
+
+      ~ContinuousHistogram.T_destination
+      ~ContinuousHistogram.call_super_init
+      ~ContinuousHistogram.dump_patches
+      ~ContinuousHistogram.training
+
+
+
+

docs/autosummary/mlcolvar.core.transform.tools.Normalization.rst

@@ -0,0 +1,44 @@
+mlcolvar.core.transform.tools.Normalization
+===========================================
+
+.. currentmodule:: mlcolvar.core.transform.tools
+
+.. autoclass:: Normalization
+   :members:                                   
+   :show-inheritance:                           
+   :inherited-members: Module,LightningModule                       
+
+
+   .. automethod:: __init__
+
+
+   .. rubric:: Methods
+
+   .. autosummary::
+
+      ~Normalization.__init__
+      ~Normalization.extra_repr
+      ~Normalization.forward
+      ~Normalization.inverse
+      ~Normalization.set_custom
+      ~Normalization.set_from_stats
+      ~Normalization.setup_from_datamodule
+
+
+
+
+..
+
+
+   .. rubric:: Attributes
+
+   .. autosummary::
+
+      ~Normalization.T_destination
+      ~Normalization.call_super_init
+      ~Normalization.dump_patches
+      ~Normalization.training
+
+
+
+