modelbase2 0.2.0__py3-none-any.whl → 0.4.0__py3-none-any.whl

modelbase2/surrogates/__init__.py

@@ -19,12 +19,11 @@ from __future__ import annotations
 import contextlib
 
 with contextlib.suppress(ImportError):
-    from ._torch import Dense, TorchSurrogate, train_torch_surrogate
+    from ._torch import TorchSurrogate, train_torch_surrogate
 
 from ._poly import PolySurrogate, train_polynomial_surrogate
 
 __all__ = [
-    "Dense",
     "PolySurrogate",
     "TorchSurrogate",
     "train_polynomial_surrogate",

modelbase2/surrogates/_poly.py

@@ -3,16 +3,27 @@ from dataclasses import dataclass
 
 import numpy as np
 import pandas as pd
-from numpy.polynomial.polynomial import Polynomial
+from numpy import polynomial
+from typing import Union, Literal
 
 from modelbase2.types import AbstractSurrogate, ArrayLike
 
-__all__ = ["PolySurrogate", "train_polynomial_surrogate"]
+__all__ = ["PolySurrogate", "PolynomialExpansion", "train_polynomial_surrogate"]
+
+# define custom type
+PolynomialExpansion = (
+    polynomial.polynomial.Polynomial
+    | polynomial.chebyshev.Chebyshev
+    | polynomial.legendre.Legendre
+    | polynomial.laguerre.Laguerre
+    | polynomial.hermite.Hermite
+    | polynomial.hermite_e.HermiteE
+)
 
 
 @dataclass(kw_only=True)
 class PolySurrogate(AbstractSurrogate):
-    model: Polynomial
+    model: PolynomialExpansion
 
     def predict_raw(self, y: np.ndarray) -> np.ndarray:
         return self.model(y)
@@ -21,15 +32,19 @@ class PolySurrogate(AbstractSurrogate):
 def train_polynomial_surrogate(
     feature: ArrayLike,
     target: ArrayLike,
+    series: Literal[
+        "Power", "Chebyshev", "Legendre", "Laguerre", "Hermite", "HermiteE"
+    ] = "Power",
     degrees: Iterable[int] = (1, 2, 3, 4, 5, 6, 7),
     surrogate_args: list[str] | None = None,
     surrogate_stoichiometries: dict[str, dict[str, float]] | None = None,
 ) -> tuple[PolySurrogate, pd.DataFrame]:
-    """Train a polynomial surrogate model.
+    """Train a surrogate model based on function series expansion.
 
     Args:
         feature: Input data as a numpy array.
         target: Output data as a numpy array.
+        series: Base functions for the surrogate model
         degrees: Degrees of the polynomial to fit to the data.
         surrogate_args: Additional arguments for the surrogate model.
         surrogate_stoichiometries: Stoichiometries for the surrogate model.
@@ -41,7 +56,19 @@ def train_polynomial_surrogate(
     feature = np.array(feature, dtype=float)
     target = np.array(target, dtype=float)
 
-    models = [Polynomial.fit(feature, target, degree) for degree in degrees]
+    # Choose numpy polynomial convenience classes
+    series_dictionary = {
+        "Power": polynomial.polynomial.Polynomial,
+        "Chebyshev": polynomial.chebyshev.Chebyshev,
+        "Legendre": polynomial.legendre.Legendre,
+        "Laguerre": polynomial.laguerre.Laguerre,
+        "Hermite": polynomial.hermite.Hermite,
+        "HermiteE": polynomial.hermite_e.HermiteE,
+    }
+
+    fn_series = series_dictionary[series]
+
+    models = [fn_series.fit(feature, target, degree) for degree in degrees]
     predictions = np.array([model(feature) for model in models], dtype=float)
     errors = np.sqrt(np.mean(np.square(predictions - target.reshape(1, -1)), axis=1))
     log_likelihood = -0.5 * np.sum(

modelbase2/surrogates/_torch.py

@@ -1,3 +1,4 @@
+from collections.abc import Callable
 from dataclasses import dataclass
 
 import numpy as np
@@ -6,12 +7,12 @@ import torch
 import tqdm
 from torch import nn
 from torch.optim.adam import Adam
+from torch.optim.optimizer import ParamsT
 
+from modelbase2.nnarchitectures import MLP, DefaultDevice
 from modelbase2.types import AbstractSurrogate
 
-__all__ = ["DefaultDevice", "Dense", "TorchSurrogate", "train_torch_surrogate"]
-
-DefaultDevice = torch.device("cpu")
+__all__ = ["TorchSurrogate", "train_torch_surrogate"]
 
 
 @dataclass(kw_only=True)
@@ -44,63 +45,6 @@ class TorchSurrogate(AbstractSurrogate):
             ).numpy()
 
 
-class Dense(nn.Module):
-    """Neural network approximator for surrogate modeling.
-
-    Attributes:
-        net: Sequential neural network model.
-
-    Methods:
-        forward: Forward pass through the neural network.
-
-    """
-
-    def __init__(self, n_inputs: int, n_outputs: int) -> None:
-        """Initializes the surrogate model with the given number of inputs and outputs.
-
-        Args:
-            n_inputs (int): The number of input features.
-            n_outputs (int): The number of output features.
-
-        Initializes a neural network with the following architecture:
-        - Linear layer with `n_inputs` inputs and 50 outputs
-        - ReLU activation
-        - Linear layer with 50 inputs and 50 outputs
-        - ReLU activation
-        - Linear layer with 50 inputs and `n_outputs` outputs
-
-        The weights of the linear layers are initialized with a normal distribution
-        (mean=0, std=0.1) and the biases are initialized to 0.
-
-        """
-        super().__init__()
-
-        self.net = nn.Sequential(
-            nn.Linear(n_inputs, 50),
-            nn.ReLU(),
-            nn.Linear(50, 50),
-            nn.ReLU(),
-            nn.Linear(50, n_outputs),
-        )
-
-        for m in self.net.modules():
-            if isinstance(m, nn.Linear):
-                nn.init.normal_(m.weight, mean=0, std=0.1)
-                nn.init.constant_(m.bias, val=0)
-
-    def forward(self, x: torch.Tensor) -> torch.Tensor:
-        """Forward pass through the neural network.
-
-        Args:
-            x: Input tensor.
-
-        Returns:
-            torch.Tensor: Output tensor.
-
-        """
-        return self.net(x)
-
-
 def _train_batched(
     aprox: nn.Module,
     features: pd.DataFrame,
@@ -182,7 +126,7 @@ def train_torch_surrogate(
     surrogate_stoichiometries: dict[str, dict[str, float]] | None = None,
     batch_size: int | None = None,
     approximator: nn.Module | None = None,
-    optimimzer_cls: type[Adam] = Adam,
+    optimimzer_cls: Callable[[ParamsT], Adam] = Adam,
     device: torch.device = DefaultDevice,
 ) -> tuple[TorchSurrogate, pd.Series]:
     """Train a PyTorch surrogate model.
@@ -205,7 +149,7 @@ def train_torch_surrogate(
         surrogate_args: List of input variable names for the surrogate model.
         surrogate_stoichiometries: Dictionary mapping reaction names to stoichiometries.
         batch_size: Size of mini-batches for training (None for full-batch).
-        approximator: Predefined neural network model (None to use default).
+        approximator: Predefined neural network model (None to use default MLP features-50-50-output).
         optimimzer_cls: Optimizer class to use for training (default: Adam).
         device: Device to run the training on (default: DefaultDevice).
 
@@ -214,9 +158,9 @@ def train_torch_surrogate(
 
     """
     if approximator is None:
-        approximator = Dense(
+        approximator = MLP(
             n_inputs=len(features.columns),
-            n_outputs=len(targets.columns),
+            layers=[50, 50, len(targets.columns)],
         ).to(device)
 
     optimizer = optimimzer_cls(approximator.parameters())

modelbase2/surrogates.py

@@ -19,6 +19,7 @@ from __future__ import annotations
 from abc import abstractmethod
 from dataclasses import dataclass
 from pathlib import Path
+from typing import TYPE_CHECKING
 
 import numpy as np
 import pandas as pd
@@ -29,6 +30,11 @@ from torch.optim.adam import Adam
 
 from modelbase2.parallel import Cache
 
+if TYPE_CHECKING:
+    from collections.abc import Callable
+
+    from torch.optim.optimizer import ParamsT
+
 __all__ = [
     "AbstractSurrogate",
     "Approximator",
@@ -251,7 +257,7 @@ def train_torch_surrogate(
     surrogate_stoichiometries: dict[str, dict[str, float]],
     batch_size: int | None = None,
     approximator: nn.Module | None = None,
-    optimimzer_cls: type[Adam] = Adam,
+    optimimzer_cls: Callable[[ParamsT], Adam] = Adam,
     device: torch.device = DefaultDevice,
 ) -> tuple[TorchSurrogate, pd.Series]:
     """Train a PyTorch surrogate model.

{modelbase2-0.2.0.dist-info → modelbase2-0.4.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: modelbase2
-Version: 0.2.0
+Version: 0.4.0
 Summary: A package to build metabolic models
 Author-email: Marvin van Aalst <marvin.vanaalst@gmail.com>
 Maintainer-email: Marvin van Aalst <marvin.vanaalst@gmail.com>
@@ -33,6 +33,7 @@ Requires-Dist: pebble>=5.0.7
 Requires-Dist: python-libsbml>=5.20.4
 Requires-Dist: scipy>=1.14.1
 Requires-Dist: seaborn>=0.13.2
+Requires-Dist: symbtools>=0.4.0
 Requires-Dist: sympy>=1.13.1
 Requires-Dist: tabulate>=0.9.0
 Requires-Dist: tqdm>=4.66.6
@@ -59,6 +60,18 @@ Description-Content-Type: text/markdown
 
 # modelbase
 
+## Installation
+
+You can install modelbase using pip: `pip install modelbase2`
+
+If you want access to the sundials solver suite via the [assimulo](https://jmodelica.org/assimulo/) package, we recommend setting up a virtual environment via [pixi](https://pixi.sh/) or [mamba / conda](https://mamba.readthedocs.io/en/latest/) using the [conda-forge](https://conda-forge.org/) channel.
+
+```bash
+pixi init
+pixi add python assimulo
+pixi add --pypi modelbase2
+```
+
 
 ## Development setup
 

{modelbase2-0.2.0.dist-info → modelbase2-0.4.0.dist-info}/RECORD

@@ -1,27 +1,31 @@
-modelbase2/__init__.py,sha256=mGA_ckCmdG8Bi1i_eiES7JShez_4qwobwRVYaMBe27E,3967
-modelbase2/distributions.py,sha256=sbqmkw3PTPK0vWOcIFqP3WXH_2Q3i-_pl_NmvAYCngM,7927
-modelbase2/fit.py,sha256=kCuwsuUs9DSRyeQXmP6RSclGGMVL1Q7XD3KLhu073yo,8010
+modelbase2/__init__.py,sha256=ArYJZoCTulkjFctJzxk7c9CDBXYRl2J9_LXr1EORilk,4048
+modelbase2/distributions.py,sha256=i456B0KYKdiGwuoVjoj3rey5lKdJmu9EVvacnfWXHGA,8744
+modelbase2/fit.py,sha256=WEI2lxLhdHFr6ax5xXrrkHUTxcEhmXyBHebfHEcXwCY,8172
 modelbase2/fns.py,sha256=8JtIzPk3DAnNHz3LoJ1ukLFTjPNO1rNCeZ7VnRmJY2o,4503
 modelbase2/label_map.py,sha256=LUwcOHQWiGfBGV5XUmPM_SOwM9IyDVcQVJ11DPfVpAo,17774
 modelbase2/linear_label_map.py,sha256=gA8CHxcehgtI6ovwZ9qNUPDvxfqbO1J1kBC_mltD4TY,10225
 modelbase2/mc.py,sha256=zlDL7e_udpIMRhSjfFJo5AwkigD0B_3H2rQxyelBuzI,16285
 modelbase2/mca.py,sha256=nMS2VnzR2VEujCFUaj9WL82CNd-oxTb3jCHP8IlJvxA,8845
-modelbase2/model.py,sha256=SJloqNi8C5oSrIyknMkXfPFDBHPW-ybrT_F9nsbREzQ,53854
-modelbase2/npe.py,sha256=fSnEyXvsS1h3S7rIbPtRW2oeJc917yt73rSw75Het3o,11199
+modelbase2/model.py,sha256=mrc2Wz1IkINX1TQvuLZFdxSINqOaZGkn8FOd40qn5BY,53842
+modelbase2/nnarchitectures.py,sha256=OA1X4UHrn7gsLuuqxK6Dhv5aiKnQflhHezYCUV-NuO8,4012
+modelbase2/npe.py,sha256=o876zHjyfJelGijSmCL0vUBfWbIhcbQyyPkwp8hZ4NA,8743
 modelbase2/parallel.py,sha256=kX4Td5YoovDwZp6kX_3cfO6QtHSS9ieJ0bMZiKs3Xv8,5002
 modelbase2/parameterise.py,sha256=7VrYxrQv0visraqUthWSnWfx-cxh2evlXbszIY5031U,690
 modelbase2/paths.py,sha256=uatKXDa79uniUB2Z3dr8eBJVuUPXDI-o_bf-DqPKq1Y,1039
-modelbase2/plot.py,sha256=tsQRUeKFKpthWOw8JqKhqcxBTBaMscJlemJzNpENAnc,22640
+modelbase2/plot.py,sha256=txKF6Xnyh2JPJ06Wu803Wn7_VijMMJ1Kbq4WQB-xKE8,22720
 modelbase2/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 modelbase2/scan.py,sha256=PvWZA4EgNS0JVMvm87bF72hSmhvo6O2KGg4vRzxve_8,18104
 modelbase2/scope.py,sha256=4twnEh8LrTmlLE-uRvubVkE3SSWejlLvtBzTCPqG3Aw,3710
-modelbase2/simulator.py,sha256=3UIAjGa6AUosd7384I4s9kVyFuY6H0bPcqIm19QpTIQ,19708
-modelbase2/surrogates.py,sha256=XGXZaFwO8CZsATbFohkuK0eQdTR_hsVEUPJyzfIKpSk,9121
+modelbase2/simulator.py,sha256=0_KXL3QUsMHtckdH4nPlzJoGNLoQaVttV7PUt2RU1m8,19823
+modelbase2/surrogates.py,sha256=Dk-YEXb-ot22W3r5Bl3rxApbCyvNdBnNCvRi_1f_NnA,9276
 modelbase2/types.py,sha256=N74pWUZDcGlSMfDkjqqwHn5dkozlqgS6Wqg-58YKdvg,9904
-modelbase2/experimental/__init__.py,sha256=En8_KLrfA5RulThEr_Zx2D1n0H0wv1L0Rj6EBwWz7RM,379
+modelbase2/experimental/__init__.py,sha256=IzgcQ7MtMII7n6cg1PMp-oZm06CNVR19ax2exvb9Ny0,444
+modelbase2/experimental/_backup.py,sha256=As-L75tMQjWfkheRPk8xcroQPKDsSSejBJ2dXqPXmVo,46644
 modelbase2/experimental/codegen.py,sha256=J_0iCtojwjmDXAfC8EqiXP0gmWaSH8MPkWvdLsZWsXU,6962
 modelbase2/experimental/diff.py,sha256=e7fjD9kqxkRUNxSevbAznd5cOlEdWJ6pj0y7Kd5KKrw,8973
 modelbase2/experimental/notes.md,sha256=YlM2biTzub6jSlx-aDZaBYsvQcGwb7NHyVAbbl2acGE,238
+modelbase2/experimental/strikepy.py,sha256=cKBs9InXR9mEPgx70Ynv0qkmAGfloksqinbpppTiC6U,19464
+modelbase2/experimental/symbolic.py,sha256=QT82TSW42RSVsvXK2WTQW1XluXnflWzrHbx6RFr_YmY,9953
 modelbase2/experimental/tex.py,sha256=M-Pdq3eQw1Huo-z1gv8EhWVO5ecJyFS8MMy9yoX81VI,13634
 modelbase2/integrators/__init__.py,sha256=kqmV6a0TRyLGR_XqbyAI652AfptYnXAUpqbSFg0CpP8,450
 modelbase2/integrators/int_assimulo.py,sha256=VEQIZFZcEovLPy8i_jR8H8XcxBRQoRVmNzzCYzInPc0,4611
@@ -33,10 +37,10 @@ modelbase2/sbml/_import.py,sha256=uT5JpFvCKjQNBFmGPba61xYShHmjzjczqnaYflilSMI,21
 modelbase2/sbml/_mathml.py,sha256=bNk9RQ_NQFDhY1R354p-gwqqHaIiyAwZ1xLPHHhiguQ,24436
 modelbase2/sbml/_name_conversion.py,sha256=XK9DEyzhrD0GBBwwjK9RA0yORrDX5c-Uvx0VtKMR5rA,1325
 modelbase2/sbml/_unit_conversion.py,sha256=dW_I6_Ou09ccwnp6LIdrPriIQnQUK5lJcjzM2Fawm6U,1927
-modelbase2/surrogates/__init__.py,sha256=0OH8BmdB5toHo7sXcZHZVASx3EEnpzcc7TeTP2tF5Ek,944
-modelbase2/surrogates/_poly.py,sha256=zKeoj3FchMAIgbnq99VkH_RfLsdy6N5lZaV0AeG9JKs,2185
-modelbase2/surrogates/_torch.py,sha256=7XHES1NyVmGTc8qmcScsfroai6rJVeV4eAGU2-oqH9A,7402
-modelbase2-0.2.0.dist-info/METADATA,sha256=f3af5XY5IQ0GkvPWYqUcnFJ79EKxxDjufw8Z2yCwk3o,2853
-modelbase2-0.2.0.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
-modelbase2-0.2.0.dist-info/licenses/LICENSE,sha256=qvG2VolmSkrcocL34V1ieOx-Rn-fpVcUbb25gHzVgZw,35079
-modelbase2-0.2.0.dist-info/RECORD,,
+modelbase2/surrogates/__init__.py,sha256=N_iXERECKvmrHiihwnyQEKOSBsmlGEuQhEotn-mWKdk,924
+modelbase2/surrogates/_poly.py,sha256=zjlNL4iYR9G51gjvZPHe3CAYF-tgACGdIBe4QUYXLQk,3110
+modelbase2/surrogates/_torch.py,sha256=CBS_3JzSgI2-xQrbq9CIXY0fJQsxbhBnWkG2TQyj7Zs,5885
+modelbase2-0.4.0.dist-info/METADATA,sha256=esuEQXsT-gP3-f_UJe65bcmXTVO5J6D-MiegzUtmxak,3344
+modelbase2-0.4.0.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
+modelbase2-0.4.0.dist-info/licenses/LICENSE,sha256=qvG2VolmSkrcocL34V1ieOx-Rn-fpVcUbb25gHzVgZw,35079
+modelbase2-0.4.0.dist-info/RECORD,,