sdg-core-lib 0.1.8.dev2__tar.gz → 0.1.8.dev4__tar.gz

{sdg_core_lib-0.1.8.dev2 → sdg_core_lib-0.1.8.dev4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: sdg-core-lib
-Version: 0.1.8.dev2
+Version: 0.1.8.dev4
 Summary: Add your description here
 Author: emiliocimino
 Author-email: emiliocimino <emilio.cimino@outlook.it>
@@ -16,8 +16,7 @@ Requires-Dist: skops==0.13.0
 Requires-Dist: statsmodels==0.14.5
 Requires-Dist: tslearn==0.7.0
 Requires-Dist: keras-tuner==1.4.8
-Requires-Dist: scipy-stubs~=1.16.2
-Requires-Dist: pandas-stubs~=2.2.3
+Requires-Dist: pydantic
 Requires-Python: >=3.12
 Description-Content-Type: text/markdown
 

{sdg_core_lib-0.1.8.dev2 → sdg_core_lib-0.1.8.dev4}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "sdg-core-lib"
-version = "0.1.8.dev2"
+version = "0.1.8.dev4"
 description = "Add your description here"
 license = "AGPL-3.0"
 readme = "README.md"
@@ -22,8 +22,7 @@ dependencies = [
     "statsmodels==0.14.5",
     "tslearn==0.7.0",
     "keras-tuner==1.4.8",
-    "scipy-stubs~=1.16.2",
-    "pandas-stubs~=2.2.3",
+    "pydantic"
 ]
 
 [dependency-groups]
@@ -38,7 +37,8 @@ dev = [
     "skops==0.13.0",
     "statsmodels==0.14.5",
     "tslearn==0.7.0",
-    "keras-tuner==1.4.8"
+    "keras-tuner==1.4.8",
+    "pydantic"
 ]
 
 test = [

{sdg_core_lib-0.1.8.dev2 → sdg_core_lib-0.1.8.dev4}/src/sdg_core_lib/data_generator/models/GANs/CTGANComponents.py

@@ -189,7 +189,6 @@ class CTGANModel(keras.Model):
         batch_size = int(batch_size)  # Convert symbolic tensor to int
         num_cats = len(self.generator.cats_disc)
         total_cond_dim = sum(self.generator.cats_disc)
-        cats_disc = tf.convert_to_tensor(self.generator.cats_disc, dtype=tf.int32)
 
         col_indices = tf.random.uniform(
             shape=[batch_size], minval=0, maxval=num_cats, dtype=tf.int32

{sdg_core_lib-0.1.8.dev2 → sdg_core_lib-0.1.8.dev4}/src/sdg_core_lib/data_generator/models/GANs/implementation/CTGAN.py

@@ -28,7 +28,7 @@ class CTGAN(UnspecializedModel):
         pac_size=10,
         learning_rate=1e-3,
         batch_size=100,
-        epochs=10,
+        epochs=50,
         gen_steps=4,
         critic_dropout=0.2,
     ):

{sdg_core_lib-0.1.8.dev2 → sdg_core_lib-0.1.8.dev4}/src/sdg_core_lib/dataset/datasets.py

@@ -2,6 +2,11 @@ from abc import ABC, abstractmethod
 import numpy as np
 
 from sdg_core_lib.dataset.columns import Numeric, Categorical, Column
+from sdg_core_lib.dataset.validation_schema import (
+    FeatureData,
+    DataSkeleton,
+    SkeletonOut,
+)
 from sdg_core_lib.preprocess.base_processor import Processor
 from sdg_core_lib.preprocess.table_processor import TableProcessor
 
@@ -75,6 +80,7 @@ class Table(Dataset):
             raise ValueError("Empty dataset")
 
         for idx, col_data in enumerate(json_data):
+            FeatureData.model_validate(col_data)
             col_type = col_data.get("column_type", "")
             col_name = col_data.get("column_name", "")
             col_values = np.array(
@@ -110,6 +116,7 @@ class Table(Dataset):
         data_map = []
 
         for col_data in sorted(skeleton, key=lambda x: int(x["column_position"])):
+            DataSkeleton.model_validate(col_data)
             col_type = col_data.get("column_type", "")
             col_name = col_data.get("column_name", "")
             col_value_type = col_data.get("column_datatype", "")
@@ -157,27 +164,27 @@ class Table(Dataset):
 
     def to_json(self) -> list[dict]:
         return [
-            {
-                "column_data": col.values.reshape(
+            FeatureData(
+                column_data=col.values.reshape(
                     -1,
                 ).tolist(),
-                "column_name": col.name,
-                "column_type": col.column_type,
-                "column_datatype": col.value_type,
-            }
+                column_name=col.name,
+                column_type=col.column_type,
+                column_datatype=col.value_type,
+            ).model_dump()
             for col in self.columns
         ]
 
     def to_skeleton(self) -> list[dict]:
         return [
-            {
-                "feature_name": col.name,
-                "feature_position": col.position,
-                "is_categorical": True if isinstance(col, Categorical) else False,
-                "type": col.value_type,
-                "feature_type": col.column_type,
-                "feature_size": str(col.get_internal_shape()[1]),
-            }
+            SkeletonOut(
+                feature_name=col.name,
+                feature_position=col.position,
+                feature_type=col.column_type,
+                type=col.value_type,
+                is_categorical=True if isinstance(col, Categorical) else False,
+                feature_size=str(col.get_internal_shape()[1]),
+            ).model_dump()
             for col in self.columns
         ]
 

sdg_core_lib-0.1.8.dev4/src/sdg_core_lib/dataset/validation_schema.py

@@ -0,0 +1,37 @@
+from pydantic import BaseModel, PositiveInt, ConfigDict
+from sdg_core_lib.commons import DataType
+from enum import Enum
+from typing import List
+
+
+class SupportedFeatureTypes(str, Enum):
+    continuous = "continuous"
+    categorical = "categorical"
+    primary_key = "primary_key"
+    group_index = "group_index"
+
+
+class BaseFeature(BaseModel):
+    model_config = ConfigDict(use_enum_values=True)
+    column_name: str
+    column_datatype: DataType
+    column_type: SupportedFeatureTypes
+
+
+class FeatureData(BaseFeature):
+    column_data: List[float | int | str] | List
+
+
+class DataSkeleton(BaseFeature):
+    column_position: int
+    column_size: PositiveInt
+
+
+class SkeletonOut(BaseModel):
+    model_config = ConfigDict(use_enum_values=True)
+    feature_name: str
+    feature_position: int
+    is_categorical: bool
+    type: DataType
+    feature_type: SupportedFeatureTypes
+    feature_size: str