pyfemtet 0.4.13__py3-none-any.whl → 0.4.14__py3-none-any.whl

pyfemtet/opt/opt/_scipy.py

@@ -15,6 +15,7 @@ from scipy.optimize import minimize, OptimizeResult
 from pyfemtet.opt._femopt_core import OptimizationStatus, generate_lhs
 from pyfemtet.opt.opt import AbstractOptimizer, logger, OptimizationMethodChecker
 from pyfemtet.core import MeshError, ModelError, SolveError
+from pyfemtet.message import Msg
 
 
 class StopIteration2(Exception):
@@ -85,11 +86,8 @@ class ScipyOptimizer(AbstractOptimizer):
         try:
             _, obj_values, cns_values = self.f(x)
         except (ModelError, MeshError, SolveError) as e:
-            logger.info(e)
-            logger.info('以下の変数で FEM 解析に失敗しました。')
-            print(self.get_parameter('dict'))
-
-            # 現状、エラーが起きたらスキップできない
+            # 現在の技術的にエラーが起きたらスキップできない
+            logger.error(Msg.ERR_FEM_FAILED_AND_CANNOT_CONTINUE)
             raise StopIteration2
 
         # constraints
@@ -134,9 +132,7 @@ class ScipyOptimizer(AbstractOptimizer):
             )
         except StopIteration2:
             res = None
-            logger.warn('Optimization has been interrupted. '
-                        'Note that you cannot acquire the OptimizationResult '
-                        'in case of `trust-constr`, `TNC`, `SLSQP` or `COBYLA`.')
+            logger.warn(Msg.WARN_INTERRUPTED_IN_SCIPY)
 
         if res is None:
             self.res = self.stop_iteration_callback.res

pyfemtet/opt/opt/_scipy_scalar.py

@@ -21,7 +21,7 @@ class ScipyScalarMethodChecker(OptimizationMethodChecker):
     def check_incomplete_bounds(self, raise_error=True): return True
 
 
-class ScipyOptimizer(AbstractOptimizer):
+class ScipyScalarOptimizer(AbstractOptimizer):
 
     def __init__(
             self,
@@ -51,10 +51,6 @@ class ScipyOptimizer(AbstractOptimizer):
         try:
             _, obj_values, cns_values = self.f(x)
         except (ModelError, MeshError, SolveError) as e:
-            logger.info(e)
-            logger.info('以下の変数で FEM 解析に失敗しました。')
-            print(self.get_parameter('dict'))
-
             # 現状、エラーが起きたらスキップできない
             raise StopIteration
 

pyfemtet/opt/prediction/base.py

@@ -0,0 +1,52 @@
+from abc import ABC, abstractmethod
+
+import numpy as np
+import pandas as pd
+
+from pyfemtet.opt._femopt_core import History
+
+
+class PredictionModelBase(ABC):
+
+    @abstractmethod
+    def fit(self, x: np.ndarray, y: np.ndarray):
+        """
+        Parameters:
+            x (np.ndarray): Input. (Point number) rows and (variable number) columns.
+            y (np.ndarray): Output. (Point number) rows and (objective number) columns.
+        """
+
+    @abstractmethod
+    def predict(self, x: np.ndarray) -> tuple[np.ndarray, np.ndarray]:
+        """
+        Parameters:
+            x (np.ndarray): Input. (Point number) rows and (variable number) columns.
+        Returns:
+            np.ndarray: (Point number) rows and (objective number) columns. Index 0 is mean and index 1 is std.
+        """
+
+
+class PyFemtetPredictionModel:
+
+    def __init__(self, history: History, df: pd.DataFrame, MetaModel: type):
+        assert issubclass(MetaModel, PredictionModelBase)
+        self.meta_model: PredictionModelBase = MetaModel()
+        self.obj_names = history.obj_names
+        self.prm_names = history.prm_names
+        self.df = df
+        self.x: np.ndarray = df[self.prm_names].values
+        self.y: np.ndarray = df[self.obj_names].values
+
+    def get_prm_index(self, prm_name):
+        return self.prm_names.index(prm_name) if prm_name in self.prm_names else None
+
+    def get_obj_index(self, obj_name):
+        return self.obj_names.index(obj_name) if obj_name in self.obj_names else None
+
+    def fit(self) -> None:
+        self.meta_model.fit(self.x, self.y)
+
+    def predict(self, x: np.ndarray) -> tuple[np.ndarray, np.ndarray]:
+        assert len(x.shape) == 2
+        assert x.shape[1] == len(self.prm_names)
+        return self.meta_model.predict(x)

pyfemtet/opt/prediction/single_task_gp.py

@@ -0,0 +1,82 @@
+import numpy as np
+import torch
+import gpytorch
+
+from botorch.models import SingleTaskGP
+from botorch.fit import fit_gpytorch_mll
+from gpytorch.mlls import ExactMarginalLogLikelihood
+
+from pyfemtet.opt.prediction.base import PredictionModelBase
+
+
+class MyStandardScaler:
+
+    # noinspection PyAttributeOutsideInit
+    def fit_transform(self, x: torch.Tensor) -> torch.Tensor:
+        self.m = x.numpy().mean(axis=0)
+        self.s = x.numpy().std(axis=0, ddof=1)
+        return self.transform(x)
+
+    def transform(self, x: torch.Tensor) -> torch.Tensor:
+        return torch.tensor((x.numpy() - self.m) / self.s).double()
+
+    def inverse_transform_mean(self, x: torch.Tensor) -> torch.Tensor:
+        return torch.tensor(x.numpy() * self.s + self.m).double()
+
+    def inverse_transform_var(self, x: torch.Tensor) -> torch.Tensor:
+        return torch.tensor(x.numpy() * self.s**2).double()
+
+
+class MyMinMaxScaler:
+
+    # noinspection PyAttributeOutsideInit
+    def fit_transform(self, x: torch.Tensor) -> torch.Tensor:
+        self.max = x.numpy().max(axis=0)
+        self.min = x.numpy().min(axis=0)
+        return self.transform(x)
+
+    def transform(self, x: torch.Tensor) -> torch.Tensor:
+        return torch.tensor((x.numpy() - self.min) / (self.max - self.min)).double()
+
+    def inverse_transform_mean(self, x: torch.Tensor) -> torch.Tensor:
+        return torch.tensor(x.numpy() * (self.max - self.min) + self.min).double()
+
+    def inverse_transform_var(self, x: torch.Tensor) -> torch.Tensor:
+        return torch.tensor(x.numpy() * (self.max - self.min)**2).double()
+
+
+class SingleTaskGPModel(PredictionModelBase):
+
+    # noinspection PyAttributeOutsideInit
+    def fit(self, x: np.ndarray, y: np.ndarray):
+        train_x = torch.tensor(x).double()
+        train_y = torch.tensor(y).double()
+
+        # Normalize the input data to the unit cube
+        self.scaler_x = MyMinMaxScaler()
+        train_x = self.scaler_x.fit_transform(train_x)
+
+        # Standardize the output data
+        self.scaler_y = MyStandardScaler()
+        train_y = self.scaler_y.fit_transform(train_y)
+
+        # Fit a Gaussian Process model using the extracted data
+        self.gp = SingleTaskGP(train_x, train_y)
+        mll = ExactMarginalLogLikelihood(self.gp.likelihood, self.gp)
+        fit_gpytorch_mll(mll)
+
+    def predict(self, x: np.ndarray) -> list[np.ndarray, np.ndarray]:
+        x = torch.tensor(x).double()
+        self.gp.eval()
+        with torch.no_grad(), gpytorch.settings.fast_pred_var():
+            # normalized
+            scaled_x = self.scaler_x.transform(x)
+            # predict
+            pred = self.gp(scaled_x)
+            scaled_mean = torch.permute(pred.mean, (1, 0))
+            scaled_var = torch.permute(pred.variance, (1, 0))
+        # unscaling
+        mean = self.scaler_y.inverse_transform_mean(scaled_mean).numpy()
+        var = self.scaler_y.inverse_transform_var(scaled_var).numpy()
+        std = np.sqrt(var)
+        return mean, std

pyfemtet/opt/visualization/complex_components/control_femtet.py

@@ -18,6 +18,7 @@ from pythoncom import com_error
 
 from pyfemtet.opt.visualization.base import PyFemtetApplicationBase, AbstractPage, logger
 from pyfemtet.opt.interface._femtet import FemtetInterface
+from pyfemtet.message import Msg
 
 
 class FemtetState(Enum):
@@ -44,7 +45,7 @@ class FemtetControl(AbstractPage):
         # button
         # noinspection PyAttributeOutsideInit
         self.connect_femtet_button = dbc.Button(
-            children='Connect to Femtet',
+            children=Msg.LABEL_CONNECT_FEMTET_BUTTON,
             id='connect-femtet-button',
             outline=True,
             color='primary',
@@ -101,7 +102,7 @@ class FemtetControl(AbstractPage):
                 # model_name is not included in femprj
                 kwargs.update(dict(model_name=None))
                 self.fem = FemtetInterface(**kwargs)
-                warning_msg = 'Analysis model name described in csv does not exist in project.'
+                warning_msg = Msg.WARN_CSV_MODEL_NAME_IS_INVALID
 
             # if 'new' femtet, Interface try to terminate Femtet when the self.fem is reconstructed.
             self.fem.quit_when_destruct = False
@@ -136,39 +137,36 @@ class FemtetControl(AbstractPage):
 
         # check holding history
         if self.application.history is None:
-            return kwargs, 'History csv is not read yet. Open your project manually.'
+            return kwargs, Msg.ERR_HISTORY_CSV_NOT_READ
 
         # get metadata
         additional_metadata = self.application.history.metadata[0]
 
         # check metadata exists
         if additional_metadata == '':
-            return kwargs, 'Cannot read project data from csv. Open your project manually.'
+            return kwargs, Msg.WARN_INVALID_METADATA
 
         # check the metadata is valid json
         try:
             d = json.loads(additional_metadata)
             femprj_path = os.path.abspath(d['femprj_path'])
         except (TypeError, json.decoder.JSONDecodeError, KeyError):
-            return kwargs, ('Cannot read project data from csv because of the invalid format. '
-                            'Valid format is like: '
-                            '"{""femprj_path"": ""c:\\path\\to\\sample.femprj"", ""model_name"": ""<model_name>""}". ' 
-                            'Open your project manually.')
+            return kwargs, Msg.WARN_INVALID_METADATA
 
         # check containing femprj
         if femprj_path is None:
-            return kwargs, 'Cannot read project data from csv. Open your project manually.'
+            return kwargs, Msg.WARN_INVALID_METADATA
 
         # check femprj exists
         if not os.path.exists(femprj_path):
-            return kwargs, '.femprj file described in csv is not found. Open your project manually.'
+            return kwargs, Msg.WARN_FEMPRJ_IN_CSV_NOT_FOUND
 
         # at this point, femprj is valid at least.
         kwargs.update({'femprj_path': femprj_path})
 
         # check model name
         model_name = d['model_name'] if 'model_name' in d.keys() else None
-        msg = '' if model_name is not None else 'Analysis model name is not specified. Open your model in the project manually.'
+        msg = '' if model_name is not None else Msg.WARN_MODEL_IN_CSV_NOT_FOUND_IN_FEMPRJ
         kwargs.update({'model_name': model_name})
         return kwargs, msg
 

pyfemtet/opt/visualization/complex_components/main_figure_creator.py

@@ -2,6 +2,7 @@ import plotly.graph_objs as go
 import plotly.express as px
 
 from pyfemtet.opt._femopt_core import History
+from pyfemtet.message import Msg
 
 
 class _ColorSet:
@@ -17,11 +18,17 @@ class _LanguageSet:
     feasible = {'label': 'feasible', True: True, False: False}
     non_domi = {'label': 'non_domi', True: True, False: False}
 
-    def __init__(self, language: str = 'ja'):
-        self.lang = language
-        if self.lang.lower() == 'ja':
-            self.feasible = {'label': '拘束条件', True: '満足', False: '違反'}
-            self.non_domi = {'label': '最適性', True: '非劣解', False: '劣解'}
+    def __init__(self):
+        self.feasible = {
+            'label': Msg.LEGEND_LABEL_CONSTRAINT,
+            True: Msg.LEGEND_LABEL_FEASIBLE,
+            False: Msg.LEGEND_LABEL_INFEASIBLE,
+        }
+        self.non_domi = {
+            'label': Msg.LEGEND_LABEL_OPTIMAL,
+            True: Msg.LEGEND_LABEL_NON_DOMI,
+            False: Msg.LEGEND_LABEL_DOMI,
+        }
 
     def localize(self, df):
         # 元のオブジェクトを変更しないようにコピー
@@ -34,7 +41,7 @@ class _LanguageSet:
         return cdf
 
 
-_ls = _LanguageSet('ja')
+_ls = _LanguageSet()
 _cs = _ColorSet()
 _ss = _SymbolSet()
 
@@ -53,7 +60,10 @@ def get_hypervolume_plot(_: History, df):
 
     fig.update_layout(
         dict(
-            title_text="ハイパーボリュームプロット",
+            title_text=Msg.GRAPH_TITLE_HYPERVOLUME,
+            transition_duration=1000,
+            xaxis_title=Msg.GRAPH_AXIS_LABEL_TRIAL,
+            yaxis_title='hypervolume',
         )
     )
 
@@ -76,7 +86,10 @@ def get_default_figure(history, df):
         fig = _get_multi_objective_pairplot(history, df)
 
     fig.update_traces(hoverinfo="none", hovertemplate=None)
-    fig.update_layout(clickmode='event+select')
+    fig.update_layout(
+        clickmode='event+select',
+        transition_duration=1000,
+    )
 
     return fig
 
@@ -120,8 +133,8 @@ def _get_single_objective_plot(history, df):
 
     fig.update_layout(
         dict(
-            title_text="目的プロット",
-            xaxis_title="解析実行回数(回)",
+            title_text=Msg.GRAPH_TITLE_SINGLE_OBJECTIVE,
+            xaxis_title=Msg.GRAPH_AXIS_LABEL_TRIAL,
             yaxis_title=obj_name,
         )
     )
@@ -182,7 +195,7 @@ def _get_multi_objective_pairplot(history, df):
 
     fig.update_layout(
         dict(
-            title_text="多目的ペアプロット",
+            title_text=Msg.GRAPH_TITLE_MULTI_OBJECTIVE,
         )
     )
 

pyfemtet/opt/visualization/complex_components/main_graph.py

@@ -22,6 +22,7 @@ import numpy as np
 
 from pyfemtet.opt.visualization.complex_components import main_figure_creator
 from pyfemtet.opt.visualization.base import PyFemtetApplicationBase, AbstractPage, logger
+from pyfemtet.message import Msg
 
 
 FLEXBOX_STYLE_ALLOW_VERTICAL_FILL = {
@@ -60,12 +61,12 @@ class MainGraph(AbstractPage):
         self.figure_creators = [
             dict(
                 tab_id='tab-objective-plot',
-                label='objectives',
+                label=Msg.TAB_LABEL_OBJECTIVES,
                 creator=main_figure_creator.get_default_figure,
             ),
             dict(
                 tab_id='tab-hypervolume-plot',
-                label='hypervolume',
+                label='Hypervolume',
                 creator=main_figure_creator.get_hypervolume_plot,
             ),
         ]
@@ -147,8 +148,6 @@ class MainGraph(AbstractPage):
             Input(self.dummy, 'children'),
             prevent_initial_call=False,)
         def redraw_main_graph(active_tab_id, _):
-            logger.debug('====================')
-            logger.debug(f'redraw_main_graph called by {callback_context.triggered_id}')
             figure, length = self.get_fig_by_tab_id(active_tab_id, with_length=True)
             return figure, length