Pr/1.4.4

.github/workflows/pr_to_pr.yml

@@ -1,9 +1,12 @@
-name: Pull request to pr/** branches
+name: Pull request and push to pr/** branches
 
 on:
  pull_request:
  branches:
  - "pr/**"
+ push:
+ branches:
+ - "pr/**"
 
 jobs:
  check-coding-style:

VERSION

@@ -1 +1 @@
-1.4.3
+1.4.4

docs/source/conf.py

@@ -24,8 +24,8 @@
 project = 'f3dasm'
 author = 'Martin van der Schelling'
 copyright = '2022, Martin van der Schelling'
-version = '1.4.3'
-release = '1.4.3'
+version = '1.4.4'
+release = '1.4.4'
 
 
 # -- General configuration ----------------------------------------------------

docs/source/rst_doc_files/classes/design/experimentsample.rst

@@ -54,6 +54,19 @@ An KeyError will be raised if the key is not found.
  >>> experiment_sample.get('param_1')
  0.0249
 
+Manually iterating over ExperimentData
+----------------------------------------
+
+The :class:`~f3dasm.design.ExperimentData` object can be manually iterated over to get :class:`~f3dasm.design.ExperimentSample` objects for each experiment:
+
+.. code-block:: python
+
+ >>> for experiment_sample in experiment_data:
+ ... print(experiment_sample)
+ ExperimentSample(0 : {'x0': 0.8184054141827567, 'x1': 0.937852542255321, 'x2': 0.7376563782762678} - {})
+ ExperimentSample(1 : {'x0': 0.7203461491873061, 'x1': 0.7320604457665572, 'x2': 0.2524387342272223} - {})
+ ExperimentSample(2 : {'x0': 0.35449352388104904, 'x1': 0.11413412225748525, 'x2': 0.1467895592274866} - {})
+
 Storing output parameters to the experiment sample
 --------------------------------------------------
 

src/f3dasm/__init__.py

@@ -37,7 +37,7 @@
 # =============================================================================
 
 
-__version__ = '1.4.3'
+__version__ = '1.4.4'
 
 
 # Log welcome message and the version of f3dasm

src/f3dasm/_src/datageneration/datagenerator.py

@@ -7,6 +7,7 @@
 
 # Standard
 import sys
+from abc import abstractmethod
 from functools import partial
 from typing import Any, Callable
 
@@ -43,30 +44,80 @@ class DataGenerator:
  """Base class for a data generator"""
 
  def pre_process(self, experiment_sample: ExperimentSample, **kwargs) -> None:
- """Function that handles the pre-processing"""
+ """Interface function that handles the pre-processing of the data generator
+
+ Notes
+ -----
+ If not implemented the function will be skipped
+
+ The experiment_sample is cached inside the data generator. This
+ allows the user to access the experiment_sample in the pre_process, execute
+ and post_process functions as a class variable called self.experiment_sample.
+ """
  ...
 
+ @abstractmethod
  def execute(self, **kwargs) -> None:
- """Function that calls the FEM simulator the pre-processing"""
- raise NotImplementedError("No execute function implemented!")
+ """Interface function that handles the execution of the data generator
+
+ Raises
+ ------
+ NotImplementedError
+ If the function is not implemented by the user
+
+ Notes
+ -----
+ The experiment_sample is cached inside the data generator. This
+ allows the user to access the experiment_sample in the pre_process, execute
+ and post_process functions as a class variable called self.experiment_sample.
+ """
+
+ ...
 
  def post_process(self, experiment_sample: ExperimentSample, **kwargs) -> None:
- """Function that handles the post-processing"""
+ """Interface function that handles the post-processing of the data generator
+
+ Notes
+ -----
+ If not implemented the function will be skipped
+
+ The experiment_sample is cached inside the data generator. This
+ allows the user to access the experiment_sample in the pre_process, execute
+ and post_process functions as a class variable called self.experiment_sample.
+ """
  ...
 
  @time_and_log
- def run(self, experiment_sample: ExperimentSample, **kwargs) -> ExperimentSample:
- """Run the data generator
+ def _run(self, experiment_sample: ExperimentSample, **kwargs) -> ExperimentSample:
+ """
+ Run the data generator
+ This function chains the following methods together
+
+ * pre_process(); to combine the experiment_sample and the parameters
+ of the data generator to an input file that can be used to run the data generator
+
+ * execute(); to run the data generator and generate the response of the experiment
+
+ * post_process(); to process the response of the experiment and store it back
+ in the experiment_sample
+
+ The function also caches the experiment_sample in the data generator. This
+ allows the user to access the experiment_sample in the pre_process, execute
+ and post_process functions as a class variable called self.experiment_sample.
 
  Parameters
  ----------
  ExperimentSample : ExperimentSample
  The design to run the data generator on
 
+ kwargs : dict
+ The keyword arguments to pass to the pre_process, execute and post_process
+
  Returns
  -------
  ExperimentSample
- Processed design
+ Processed design with the response of the data generator saved in the
+ experiment_sample
  """
  # Cache the design
  self.experiment_sample: ExperimentSample = experiment_sample
@@ -88,7 +139,25 @@ def _post_simulation(self) -> None:
  ...
 
  def add_pre_process(self, func: Callable, **kwargs):
+ """Add a pre-processing function to the data generator
+
+ Parameters
+ ----------
+ func : Callable
+ The function to add to the pre-processing
+ kwargs : dict
+ The keyword arguments to pass to the pre-processing function
+ """
  self.pre_process = partial(func, **kwargs)
 
  def add_post_process(self, func: Callable, **kwargs):
+ """Add a post-processing function to the data generator
+
+ Parameters
+ ----------
+ func : Callable
+ The function to add to the post-processing
+ kwargs : dict
+ The keyword arguments to pass to the post-processing function
+ """
  self.post_process = partial(func, **kwargs)

src/f3dasm/_src/datageneration/functions/function.py

@@ -99,7 +99,7 @@ def execute(self, experiment_sample: ExperimentSample) -> ExperimentSample:
  experiment_sample["y"] = self(x).ravel().astype(np.float32)
  return experiment_sample
 
- def run(self, experiment_sample: ExperimentSample, **kwargs) -> ExperimentSample:
+ def _run(self, experiment_sample: ExperimentSample, **kwargs) -> ExperimentSample:
  return self.execute(experiment_sample)
 
  def _retrieve_original_input(self, x: np.ndarray):

src/f3dasm/_src/design/domain.py

@@ -8,10 +8,11 @@
 from __future__ import annotations
 
 # Standard
+import math
 import pickle
 from dataclasses import dataclass, field
 from pathlib import Path
-from typing import Any, Dict, Iterator, List, Sequence, Type
+from typing import Any, Dict, Iterable, Iterator, List, Sequence, Type
 
 # Third-party core
 import numpy as np
@@ -252,7 +253,14 @@ def add_int(self, name: str, low: int, high: int, step: int = 1):
  >>> domain.add_int('param1', 0, 10, 2)
  >>> domain.space
  {'param1': DiscreteParameter(lower_bound=0, upper_bound=10, step=2)}
+
+ Note
+ ----
+ If the lower and upper bound are equal, then a constant parameter
+ will be added to the domain!
  """
+ if low == high:
+ self.add_constant(name, low)
  self._add(name, DiscreteParameter(low, high, step))
 
  def add_float(self, name: str, low: float, high: float, log: bool = False):
@@ -275,8 +283,16 @@ def add_float(self, name: str, low: float, high: float, log: bool = False):
  >>> domain.add_float('param1', 0., 10., log=True)
  >>> domain.space
  {'param1': ContinuousParameter(lower_bound=0., upper_bound=10., log=True)}
+
+ Note
+ ----
+ If the lower and upper bound are equal, then a constant parameter
+ will be added to the domain!
  """
- self._add(name, ContinuousParameter(low, high, log))
+ if math.isclose(low, high):
+ self.add_constant(name, low)
+ else:
+ self._add(name, ContinuousParameter(low, high, log))
 
  def add_category(self, name: str, categories: Sequence[CategoricalType]):
  """Add a new categorical input parameter to the domain.
@@ -573,6 +589,38 @@ def _filter(self, type: Type[Parameter]) -> Domain:
  if isinstance(parameter, type)}
  )
 
+ def select(self, names: str | Iterable[str]) -> Domain:
+ """Select a subset of parameters from the domain.
+
+ Parameters
+ ----------
+
+ names : str or Iterable[str]
+ The names of the parameters to select.
+
+ Returns
+ -------
+ Domain
+ A new domain with the selected parameters.
+
+ Example
+ -------
+ >>> domain = Domain()
+ >>> domain.space = {
+ ... 'param1': ContinuousParameter(lower_bound=0., upper_bound=1.),
+ ... 'param2': DiscreteParameter(lower_bound=0, upper_bound=8),
+ ... 'param3': CategoricalParameter(categories=['cat1', 'cat2'])
+ ... }
+ >>> domain.select(['param1', 'param3'])
+ Domain({'param1': ContinuousParameter(lower_bound=0, upper_bound=1),
+ 'param3': CategoricalParameter(categories=['cat1', 'cat2'])})
+ """
+
+ if isinstance(names, str):
+ names = [names]
+
+ return Domain(space={key: self.space[key] for key in names})
+
 # Miscellaneous
 # =============================================================================
 

src/f3dasm/_src/experimentdata/_data.py

@@ -291,6 +291,23 @@ def n_best_samples(self, nosamples: int, column_name: List[str] | str) -> pd.Dat
  """
  return self.data.nsmallest(n=nosamples, columns=column_name)
 
+ def select_columns(self, columns: Iterable[str] | str) -> _Data:
+ """Filter the data on the selected columns.
+
+ Parameters
+ ----------
+ columns : Iterable[str] | str
+ The columns to select.
+
+ Returns
+ -------
+ _Data
+ The data only with the selected columns
+ """
+ # This is necessary otherwise self.data[columns] will be a Series
+ if isinstance(columns, str):
+ columns = [columns]
+ return _Data(self.data[columns])
 # Append and remove data
 # =============================================================================