Model Repository Parser for UML Diagrams (#144)

* Model Repository Parser for UML Diagrams

fedot_ind/tools/uml/uml.py

@@ -0,0 +1,48 @@
+import re
+import os
+
+from fedot_ind.api.utils.path_lib import DEFAULT_PATH_MODELS
+
+with open(DEFAULT_PATH_MODELS, 'r') as file:
+    file_content = [string for string in file.read().split("\n") if not string.startswith("from")]
+file_content = file_content[:next((i for i, line in enumerate(file_content) if line.startswith("def")), None)]
+file_content = file_content[next((i for i, line in enumerate(file_content) if line == '}'), None) + 1:]
+
+pattern = re.compile(r'\b[A-Z_]+\b')
+
+substrings = []
+for s in file_content:
+    matches = pattern.findall(s)
+    substrings.extend(matches)
+substrings = list(set(substrings))
+
+res = ["@startuml", "'https://plantuml.com/sequence-diagram", "", "class AtomizedModel {", "    Enum", "}", ""]
+for ind in substrings:
+    res += ["abstract " + ind]
+res += ["", ""]
+
+res += ["INDUSTRIAL_CLF_PREPROC_MODEL --> AtomizedModel",
+        "AtomizedModel <-- SKLEARN_CLF_MODELS",
+        "FEDOT_PREPROC_MODEL --> AtomizedModel",
+        "AtomizedModel <- INDUSTRIAL_PREPROC_MODEL",
+        "SKLEARN_REG_MODELS --> AtomizedModel",
+        "AtomizedModel <-- FORECASTING_MODELS",
+        "FORECASTING_PREPROC -> AtomizedModel",
+        "NEURAL_MODEL -> AtomizedModel", ""]
+
+for ind in substrings:
+    start_index = next((i for i, line in enumerate(file_content) if line.endswith(ind + " = {")), None)
+    end_index = next((i for i, line in enumerate(file_content[start_index:]) if line.endswith("}")),
+                     None) + 1 + start_index
+    list_of_strings = file_content[start_index:end_index]
+    pattern = re.compile(r"': ([^']*)")
+    formatted_strings = [pattern.search(string).group(1) if pattern.search(string) else ""
+                         for string in list_of_strings]
+    res = res + ["abstract " + ind + " {"] + [string for string in formatted_strings if
+                                              string.strip() and string.strip() != "{"] + ["}", ""]
+
+with open('your_file.puml', 'w') as f:
+    for line in res + ["@enduml"]:
+        f.write(f"{line}\n")
+
+os.system('python -m plantuml your_file.puml')

fedot_ind/tools/uml/your_file.puml

@@ -0,0 +1,101 @@
+@startuml
+'https://plantuml.com/sequence-diagram
+
+class AtomizedModel {
+    Enum
+}
+
+abstract FEDOT_PREPROC_MODEL
+abstract NEURAL_MODEL
+abstract INDUSTRIAL_PREPROC_MODEL
+abstract SKLEARN_CLF_MODELS
+abstract INDUSTRIAL_CLF_PREPROC_MODEL
+abstract SKLEARN_REG_MODELS
+abstract FORECASTING_PREPROC
+abstract FORECASTING_MODELS
+
+
+INDUSTRIAL_CLF_PREPROC_MODEL --> AtomizedModel
+AtomizedModel <-- SKLEARN_CLF_MODELS
+FEDOT_PREPROC_MODEL --> AtomizedModel
+AtomizedModel <- INDUSTRIAL_PREPROC_MODEL
+SKLEARN_REG_MODELS --> AtomizedModel
+AtomizedModel <-- FORECASTING_MODELS
+FORECASTING_PREPROC -> AtomizedModel
+NEURAL_MODEL -> AtomizedModel
+
+abstract FEDOT_PREPROC_MODEL {
+ScalingImplementation,
+NormalizationImplementation,
+ImputationImplementation,
+KernelPCAImplementation,
+TopologicalFeaturesImplementation
+}
+
+abstract NEURAL_MODEL {
+InceptionTimeModel,
+OmniScaleModel,
+ResNetModel,
+NBeatsModel,
+TSTModel,
+XCModel,
+DummyOverComplicatedNeuralNetwork,
+LoraModel
+}
+
+abstract INDUSTRIAL_PREPROC_MODEL {
+ChannelCentroidFilter,
+EigenBasisImplementation,
+WaveletBasisImplementation,
+FourierBasisImplementation,
+RecurrenceExtractor,
+QuantileExtractor,
+RiemannExtractor,
+TopologicalFeaturesImplementation,
+MiniRocketExtractor,
+ChronosExtractor,
+IsolationForestClassImplementation,
+IsolationForestRegImplementation,
+}
+
+abstract SKLEARN_CLF_MODELS {
+GradientBoostingClassifier,
+SklearnLogReg,
+DecisionTreeClassifier,
+RandomForestClassifier,
+MLPClassifier,
+LGBMClassifier
+}
+
+abstract INDUSTRIAL_CLF_PREPROC_MODEL {
+DecomposerClassImplementation,
+ResampleImplementation,
+}
+
+abstract SKLEARN_REG_MODELS {
+XGBRegressor,
+SklearnSGD,
+ExtraTreesRegressor,
+SklearnRidgeReg,
+SklearnLassoReg,
+DecisionTreeRegressor,
+LGBMRegressor,
+}
+
+abstract FORECASTING_PREPROC {
+LaggedTransformationImplementation,
+SparseLaggedTransformationImplementation,
+TsSmoothingImplementation,
+GaussianFilterImplementation,
+ExogDataTransformationImplementation,
+}
+
+abstract FORECASTING_MODELS {
+AutoRegImplementation,
+STLForecastARIMAImplementation,
+ExpSmoothingImplementation,
+CGRUImplementation,
+GLMIndustrial
+}
+
+@enduml