validmind 2.12.2__tar.gz → 2.12.4__tar.gz

{validmind-2.12.2 → validmind-2.12.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: validmind
-Version: 2.12.2
+Version: 2.12.4
 Summary: ValidMind Library
 License: DUAL LICENSE NOTICE
          
@@ -692,13 +692,15 @@ License: DUAL LICENSE NOTICE
 License-File: LICENSE
 Author: Andres Rodriguez
 Author-email: andres@validmind.ai
-Requires-Python: >=3.9,<3.13
+Requires-Python: >=3.9,<3.15
 Classifier: License :: Other/Proprietary License
 Classifier: Programming Language :: Python :: 3
 Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Programming Language :: Python :: 3.14
 Provides-Extra: all
 Provides-Extra: credit-risk
 Provides-Extra: datasets
@@ -712,17 +714,28 @@ Provides-Extra: stats
 Provides-Extra: xgboost
 Requires-Dist: aiohttp[speedups] (<3.13.1)
 Requires-Dist: anywidget
-Requires-Dist: arch ; extra == "all"
-Requires-Dist: arch ; extra == "stats"
+Requires-Dist: arch (>=7.0.0) ; extra == "all"
+Requires-Dist: arch (>=7.0.0) ; extra == "stats"
 Requires-Dist: beautifulsoup4
 Requires-Dist: bert-score (>=0.3.13) ; extra == "all"
 Requires-Dist: bert-score (>=0.3.13) ; extra == "nlp"
-Requires-Dist: datasets (>=2.10.0,<3.0.0) ; extra == "datasets"
-Requires-Dist: deepeval (>=3.7.0) ; extra == "llm"
+Requires-Dist: datasets (>=2.10.0,<3.0.0) ; (python_version < "3.14") and (extra == "all")
+Requires-Dist: datasets (>=2.10.0,<3.0.0) ; (python_version < "3.14") and (extra == "datasets")
+Requires-Dist: datasets (>=2.10.0,<3.0.0) ; (python_version < "3.14") and (extra == "llm")
+Requires-Dist: datasets (>=2.10.0,<3.0.0) ; (python_version < "3.14") and (extra == "nlp")
+Requires-Dist: datasets (>=4.4,<5.0.0) ; (python_version >= "3.14") and (extra == "all")
+Requires-Dist: datasets (>=4.4,<5.0.0) ; (python_version >= "3.14") and (extra == "datasets")
+Requires-Dist: datasets (>=4.4,<5.0.0) ; (python_version >= "3.14") and (extra == "llm")
+Requires-Dist: datasets (>=4.4,<5.0.0) ; (python_version >= "3.14") and (extra == "nlp")
+Requires-Dist: deepeval (>=3.7.0,<=3.7.5) ; extra == "llm"
 Requires-Dist: evaluate ; extra == "all"
 Requires-Dist: evaluate ; extra == "nlp"
 Requires-Dist: ipywidgets
 Requires-Dist: kaleido (>=0.2.1,!=0.2.1.post1,<1.0.0)
+Requires-Dist: langchain (>=0.2.0,<0.4.0) ; extra == "all"
+Requires-Dist: langchain (>=0.2.0,<0.4.0) ; extra == "llm"
+Requires-Dist: langchain-community (>=0.2.0,<0.4.0) ; extra == "all"
+Requires-Dist: langchain-community (>=0.2.0,<0.4.0) ; extra == "llm"
 Requires-Dist: langchain-openai (>=0.1.8) ; extra == "all"
 Requires-Dist: langchain-openai (>=0.1.8) ; extra == "llm"
 Requires-Dist: langdetect ; extra == "all"
@@ -732,14 +745,27 @@ Requires-Dist: mistune (>=3.0.2,<4.0.0)
 Requires-Dist: nest-asyncio (>=1.6.0,<2.0.0)
 Requires-Dist: nltk (>=3.8.1,<4.0.0) ; extra == "all"
 Requires-Dist: nltk (>=3.8.1,<4.0.0) ; extra == "nlp"
-Requires-Dist: numpy (>=1.23,<2.0.0)
+Requires-Dist: numpy (>=1.23,<2.0.0) ; python_version < "3.14"
+Requires-Dist: numpy (>=2.3,<3.0.0) ; python_version >= "3.14"
 Requires-Dist: openai (>=1)
 Requires-Dist: pandas (>=2.0.3,<3.0.0)
 Requires-Dist: plotly (>=6.0.0)
 Requires-Dist: polars
-Requires-Dist: presidio-analyzer ; extra == "pii-detection"
+Requires-Dist: presidio-analyzer (<2.2.360) ; (python_version >= "3.14") and (extra == "pii-detection")
+Requires-Dist: presidio-analyzer ; (python_version < "3.14") and (extra == "pii-detection")
 Requires-Dist: presidio-structured ; extra == "pii-detection"
-Requires-Dist: pyarrow (<16) ; extra == "nlp"
+Requires-Dist: pyarrow (<16) ; (python_version < "3.13") and (extra == "all")
+Requires-Dist: pyarrow (<16) ; (python_version < "3.13") and (extra == "datasets")
+Requires-Dist: pyarrow (<16) ; (python_version < "3.13") and (extra == "llm")
+Requires-Dist: pyarrow (<16) ; (python_version < "3.13") and (extra == "nlp")
+Requires-Dist: pyarrow (>=20,<21) ; (python_version == "3.13") and (extra == "all")
+Requires-Dist: pyarrow (>=20,<21) ; (python_version == "3.13") and (extra == "datasets")
+Requires-Dist: pyarrow (>=20,<21) ; (python_version == "3.13") and (extra == "llm")
+Requires-Dist: pyarrow (>=20,<21) ; (python_version == "3.13") and (extra == "nlp")
+Requires-Dist: pyarrow (>=22,<23) ; (python_version >= "3.14") and (extra == "all")
+Requires-Dist: pyarrow (>=22,<23) ; (python_version >= "3.14") and (extra == "datasets")
+Requires-Dist: pyarrow (>=22,<23) ; (python_version >= "3.14") and (extra == "llm")
+Requires-Dist: pyarrow (>=22,<23) ; (python_version >= "3.14") and (extra == "nlp")
 Requires-Dist: pycocoevalcap ; extra == "all"
 Requires-Dist: pycocoevalcap ; extra == "llm"
 Requires-Dist: python-dotenv
@@ -756,6 +782,8 @@ Requires-Dist: seaborn
 Requires-Dist: sentencepiece (>=0.2.0,<0.3.0) ; extra == "all"
 Requires-Dist: sentencepiece (>=0.2.0,<0.3.0) ; extra == "huggingface"
 Requires-Dist: sentencepiece (>=0.2.0,<0.3.0) ; extra == "llm"
+Requires-Dist: setuptools (<82) ; extra == "all"
+Requires-Dist: setuptools (<82) ; extra == "credit-risk"
 Requires-Dist: shap (>=0.46.0) ; extra == "all"
 Requires-Dist: shap (>=0.46.0) ; extra == "explainability"
 Requires-Dist: statsmodels (>=0.14.2,<0.15.0) ; extra == "all"
@@ -771,6 +799,7 @@ Requires-Dist: tqdm
 Requires-Dist: transformers (>=4.32.0,<5.0.0) ; extra == "all"
 Requires-Dist: transformers (>=4.32.0,<5.0.0) ; extra == "huggingface"
 Requires-Dist: transformers (>=4.32.0,<5.0.0) ; extra == "llm"
+Requires-Dist: transformers (>=4.32.0,<5.0.0) ; extra == "nlp"
 Requires-Dist: xgboost (>=1.5.2,<3) ; extra == "all"
 Requires-Dist: xgboost (>=1.5.2,<3) ; extra == "xgboost"
 Description-Content-Type: text/markdown

{validmind-2.12.2 → validmind-2.12.4}/pyproject.toml

@@ -1,9 +1,9 @@
 [project]
 name = "validmind"
-version = "2.12.2"
+version = "2.12.4"
 description = "ValidMind Library"
 readme = "README.pypi.md"
-requires-python = ">=3.9,<3.13"
+requires-python = ">=3.9,<3.15"
 license = { file = "LICENSE" }
 authors = [
   { name = "Andres Rodriguez", email = "andres@validmind.ai" },
@@ -19,7 +19,8 @@ dependencies = [
   "matplotlib",
   "mistune (>=3.0.2,<4.0.0)",
   "nest-asyncio (>=1.6.0,<2.0.0)",
-  "numpy (>=1.23,<2.0.0)",
+  "numpy (>=1.23,<2.0.0) ; python_version < '3.14'",
+  "numpy (>=2.3,<3.0.0) ; python_version >= '3.14'",
   "openai (>=1)",
   "pandas (>=2.0.3,<3.0.0)",
   "plotly (>=6.0.0)",
@@ -41,6 +42,8 @@ all = [
   "transformers (>=4.32.0,<5.0.0)",
   "pycocoevalcap",
   "ragas (>=0.2.3,<=0.2.7)",
+  "langchain (>=0.2.0,<0.4.0)",
+  "langchain-community (>=0.2.0,<0.4.0)",
   "sentencepiece (>=0.2.0,<0.3.0)",
   "langchain-openai (>=0.1.8)",
   "scipy",
@@ -48,12 +51,18 @@ all = [
   "langdetect",
   "nltk (>=3.8.1,<4.0.0)",
   "textblob (>=0.18.0.post0,<0.19.0)",
+  "datasets (>=2.10.0,<3.0.0) ; python_version < '3.14'",
+  "datasets (>=4.4,<5.0.0) ; python_version >= '3.14'",
   "evaluate",
   "rouge (>=1)",
   "bert-score (>=0.3.13)",
-  "arch",
+  "pyarrow (<16) ; python_version < '3.13'",
+  "pyarrow (>=20,<21) ; python_version == '3.13'",
+  "pyarrow (>=22,<23) ; python_version >= '3.14'",
+  "arch (>=7.0.0)",
   "shap (>=0.46.0)",
   "scorecardpy==0.1.9.6",
+  "setuptools (<82)",
 ]
 huggingface = [
   "transformers (>=4.32.0,<5.0.0)",
@@ -64,26 +73,48 @@ llm = [
   "transformers (>=4.32.0,<5.0.0)",
   "pycocoevalcap",
   "ragas (>=0.2.3,<=0.2.7)",
+  "langchain (>=0.2.0,<0.4.0)",
+  "langchain-community (>=0.2.0,<0.4.0)",
+  "datasets (>=2.10.0,<3.0.0) ; python_version < '3.14'",
+  "datasets (>=4.4,<5.0.0) ; python_version >= '3.14'",
   "sentencepiece (>=0.2.0,<0.3.0)",
   "langchain-openai (>=0.1.8)",
-  "deepeval (>=3.7.0)",
+  "pyarrow (<16) ; python_version < '3.13'",
+  "pyarrow (>=20,<21) ; python_version == '3.13'",
+  "pyarrow (>=22,<23) ; python_version >= '3.14'",
+  "deepeval (>=3.7.0,<=3.7.5)",
 ]
 nlp = [
   "langdetect",
   "nltk (>=3.8.1,<4.0.0)",
   "textblob (>=0.18.0.post0,<0.19.0)",
+  "datasets (>=2.10.0,<3.0.0) ; python_version < '3.14'",
+  "datasets (>=4.4,<5.0.0) ; python_version >= '3.14'",
+  "transformers (>=4.32.0,<5.0.0)",
   "evaluate",
   "rouge (>=1)",
   "bert-score (>=0.3.13)",
-  "pyarrow (<16)",
+  "pyarrow (<16) ; python_version < '3.13'",
+  "pyarrow (>=20,<21) ; python_version == '3.13'",
+  "pyarrow (>=22,<23) ; python_version >= '3.14'",
 ]
 pytorch = ["torch (>=2.0.0)"]
-stats = ["scipy", "statsmodels (>=0.14.2,<0.15.0)", "arch"]
+stats = ["scipy", "statsmodels (>=0.14.2,<0.15.0)", "arch (>=7.0.0)"]
 xgboost = ["xgboost (>=1.5.2,<3)"]
 explainability = ["shap (>=0.46.0)"]
-credit_risk = ["scorecardpy==0.1.9.6"]
-datasets = ["datasets (>=2.10.0,<3.0.0)"]
-pii-detection = ["presidio-analyzer", "presidio-structured"]
+credit_risk = ["scorecardpy==0.1.9.6", "setuptools (<82)"]
+datasets = [
+  "datasets (>=2.10.0,<3.0.0) ; python_version < '3.14'",
+  "datasets (>=4.4,<5.0.0) ; python_version >= '3.14'",
+  "pyarrow (<16) ; python_version < '3.13'",
+  "pyarrow (>=20,<21) ; python_version == '3.13'",
+  "pyarrow (>=22,<23) ; python_version >= '3.14'",
+]
+pii-detection = [
+  "presidio-analyzer ; python_version < '3.14'",
+  "presidio-analyzer (<2.2.360) ; python_version >= '3.14'",
+  "presidio-structured",
+]
 
 [tool.poetry.group.dev.dependencies]
 black = "^22.1.0"

validmind-2.12.4/validmind/__version__.py

@@ -0,0 +1 @@
+__version__ = "2.12.4"

{validmind-2.12.2 → validmind-2.12.4}/validmind/client.py

@@ -282,33 +282,23 @@ def init_r_model(
     """
     Initialize a VM Model from an R model.
 
-    LogisticRegression and LinearRegression models are converted to sklearn models by extracting
-    the coefficients and intercept from the R model. XGB models are loaded using the xgboost
-    since xgb models saved in .json or .bin format can be loaded directly with either Python or R.
+    The model must first be saved to an .RData file using the R package's
+    ``save_model()`` function. This function then uses rpy2 to load the model
+    into Python for testing and validation.
 
     Args:
-        model_path (str): The path to the R model saved as an RDS or XGB file.
+        model_path (str): The path to the R model saved as an .RData file.
         input_id (str): The input ID for the model. Defaults to "model".
 
     Returns:
         VMModel: A VM Model instance.
     """
-
-    # TODO: proper check for supported models
-    #
-    # if model.get("method") not in R_MODEL_METHODS:
-    #     raise UnsupportedRModelError(
-    #         "R model method must be one of {}. Got {}".format(
-    #             R_MODEL_METHODS, model.get("method")
-    #         )
-    #     )
-
-    # first we need to load the model using rpy2
-    # since rpy2 is an extra we need to conditionally import it
     try:
         import rpy2.robjects as robjects
-    except ImportError:
-        raise MissingRExtrasError()
+    except Exception as e:
+        raise MissingRExtrasError(
+            f"`rpy2` is required to use R models. Import failed: {e}"
+        )
 
     r = robjects.r
     loaded_objects = r.load(model_path)
@@ -321,6 +311,14 @@ def init_r_model(
         input_id=input_id,
     )
 
+    metadata = get_model_info(vm_model)
+    log_input(
+        input_id=input_id,
+        type="model",
+        metadata=metadata,
+    )
+    input_registry.add(key=input_id, obj=vm_model)
+
     return vm_model
 
 

{validmind-2.12.2 → validmind-2.12.4}/validmind/models/r_model.py

@@ -98,11 +98,10 @@ class RModel(VMModel):
         Instead, there is a global predict() method that returns the predicted
         values according to the model type.
         """
-        # Use the predict method on the loaded model (assuming the model's name in R is 'model')
         predicted_probs = self.r.predict(
             self.model, newdata=new_data_r, type="response"
         )
-        return predicted_probs
+        return np.array(predicted_probs)
 
     def r_xgb_predict(self, new_data_r):
         """
@@ -114,7 +113,7 @@ class RModel(VMModel):
         predicted_probs = self.r.predict(
             self.model, newdata=new_data_r, type="response"
         )
-        return predicted_probs
+        return np.array(predicted_probs)
 
     def predict_proba(self, new_data):
         """
@@ -127,24 +126,30 @@ class RModel(VMModel):
         Converts the predicted probabilities to classes
         """
         try:
+            from rpy2.robjects import conversion, default_converter
             from rpy2.robjects import pandas2ri
         except ImportError:
             raise MissingRExtrasError()
 
-        # Activate the pandas conversion for rpy2
-        pandas2ri.activate()
-
         new_data_class = get_full_class_name(new_data)
 
         if new_data_class == "numpy.ndarray":
-            # We need to reconstruct the DataFrame from the ndarray using the column names
-            new_data = pd.DataFrame(new_data, columns=self.test_ds.feature_columns)
+            # Reconstruct a DataFrame from the ndarray using column names
+            # from the model's training data
+            try:
+                model_terms = list(self.r.attr(self.model.rx2["terms"], "term.labels"))
+                new_data = pd.DataFrame(new_data, columns=model_terms)
+            except Exception:
+                # Fallback: use generic column names
+                new_data = pd.DataFrame(new_data)
         elif new_data_class != "pandas.core.frame.DataFrame":
             raise ValueError(
                 f"new_data must be a DataFrame or ndarray. Got {new_data_class}"
             )
 
-        new_data_r = pandas2ri.py2rpy(new_data)
+        # Use context manager for pandas conversion (activate/deactivate is deprecated)
+        with conversion.localconverter(default_converter + pandas2ri.converter):
+            new_data_r = conversion.get_conversion().py2rpy(new_data)
 
         if self.__model_class() == "xgb.Booster":
             predicted_probs = self.r_xgb_predict(new_data_r)

{validmind-2.12.2 → validmind-2.12.4}/validmind/template.py

@@ -190,20 +190,37 @@ def _create_section_html(tree: List[Dict[str, Any]]) -> str:
     return StatefulHTMLRenderer.render_accordion(accordion_items, accordion_titles)
 
 
+def _print_section_tree(sections: List[Dict[str, Any]], indent: int = 0) -> None:
+    """Print a plain-text representation of the template section tree."""
+    prefix = "  " * indent
+    for i, section in enumerate(sections):
+        number = f"{i + 1}." if indent == 0 else ""
+        print(f"{prefix}{number} {section['title']} ('{section['id']}')")
+
+        for content in section.get("contents", []):
+            content_type = CONTENT_TYPE_MAP.get(
+                content["content_type"], content["content_type"]
+            )
+            print(f"{prefix}  - [{content_type}] {content['content_id']}")
+
+        if section.get("sections"):
+            _print_section_tree(section["sections"], indent + 1)
+
+
 def preview_template(template: str) -> None:
-    """Preview a template in Jupyter Notebook.
+    """Preview a template in Jupyter Notebook or plain text.
 
     Args:
         template (dict): The template to preview.
     """
+    section_tree = _convert_sections_to_section_tree(template["sections"])
+
     if not is_notebook():
-        logger.warning("preview_template() only works in Jupyter Notebook")
+        _print_section_tree(section_tree)
         return
 
     html_content = StatefulHTMLRenderer.get_base_css()
-    html_content += _create_section_html(
-        _convert_sections_to_section_tree(template["sections"])
-    )
+    html_content += _create_section_html(section_tree)
     display(html_content)
 
 

{validmind-2.12.2 → validmind-2.12.4}/validmind/utils.py

@@ -359,6 +359,9 @@ def format_number(number):
 
 def format_dataframe(df: pd.DataFrame) -> pd.DataFrame:
     """Format a pandas DataFrame for display purposes."""
+    if not is_notebook():
+        return df
+
     df = df.style.set_properties(**{"text-align": "left"}).hide(axis="index")
     return df.set_table_styles([dict(selector="th", props=[("text-align", "left")])])
 
@@ -509,7 +512,7 @@ def get_dataset_info(dataset):
 
 
 def preview_test_config(config):
-    """Preview test configuration in a collapsible HTML section.
+    """Preview test configuration in a collapsible HTML section or plain text.
 
     Args:
         config (dict): Test configuration dictionary.
@@ -521,6 +524,10 @@ def preview_test_config(config):
         logger.error(f"JSON serialization failed: {e}")
         return
 
+    if not is_notebook():
+        print(formatted_json)
+        return
+
     collapsible_html = f"""
     <script>
     function toggleOutput() {{
@@ -537,6 +544,9 @@ def preview_test_config(config):
 
 def display(widget_or_html, syntax_highlighting=True, mathjax=True):
     """Display HTML content with extra goodies (syntax highlighting, MathJax, etc.)."""
+    if not is_notebook():
+        return
+
     if hasattr(widget_or_html, "to_html"):
         html_content = widget_or_html.to_html()
         ipy_display(HTML(html_content))

{validmind-2.12.2 → validmind-2.12.4}/validmind/vm_models/html_progress.py

@@ -10,6 +10,7 @@ import uuid
 
 from IPython.display import HTML, display, update_display
 
+from ..utils import is_notebook
 from .html_renderer import StatefulHTMLRenderer
 
 
@@ -33,6 +34,10 @@ class HTMLProgressBar:
     def display(self):
         """Display the progress bar."""
         if not self._displayed:
+            if not is_notebook():
+                self._displayed = True
+                return
+
             html_content = StatefulHTMLRenderer.render_live_progress_bar(
                 max_value=self.max_value,
                 description=self.description,
@@ -57,6 +62,9 @@ class HTMLProgressBar:
 
     def _update_fallback(self):
         """Fallback method to update progress bar by replacing the entire HTML."""
+        if not is_notebook():
+            return
+
         html_content = StatefulHTMLRenderer.render_progress_bar(
             value=self.value,
             max_value=self.max_value,
@@ -74,7 +82,7 @@ class HTMLProgressBar:
 
     def close(self):
         """Close/hide the progress bar."""
-        if self._displayed:
+        if self._displayed and is_notebook():
             final_html = StatefulHTMLRenderer.render_progress_bar(
                 value=self.value,
                 max_value=self.max_value,
@@ -101,13 +109,17 @@ class HTMLLabel:
     def display(self):
         """Display the label."""
         if not self._displayed:
+            self._displayed = True
+
+            if not is_notebook():
+                return
+
             html_content = f"""
             <div class="vm-label" id="{self.label_id}" style="font-weight: bold; margin: 5px 0;">
                 {self.value}
             </div>
             """
             display(HTML(html_content), display_id=self._display_id)
-            self._displayed = True
 
     def update(self, value: str):
         """Update the label text.
@@ -117,7 +129,7 @@ class HTMLLabel:
         """
         self.value = value
 
-        if self._displayed:
+        if self._displayed and is_notebook():
             update_script = f"""
             <script>
             var labelElement = document.getElementById('{self.label_id}');
@@ -152,10 +164,17 @@ class HTMLBox:
     def display(self):
         """Display the box and its children."""
         if not self._displayed:
-            child_html_parts = []
+            self._displayed = True
+
             for child in self.children:
                 if hasattr(child, "display"):
                     child.display()
+
+            if not is_notebook():
+                return
+
+            child_html_parts = []
+            for child in self.children:
                 if hasattr(child, "bar_id"):
                     child_html_parts.append(f'<div id="{child.bar_id}"></div>')
                 elif hasattr(child, "label_id"):
@@ -167,4 +186,3 @@ class HTMLBox:
             </div>
             """
             display(HTML(html_content), display_id=self._display_id)
-            self._displayed = True

validmind-2.12.2/validmind/__version__.py

@@ -1 +0,0 @@
-__version__ = "2.12.2"