validmind 2.7.10__tar.gz → 2.7.12__tar.gz

{validmind-2.7.10 → validmind-2.7.12}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: validmind
-Version: 2.7.10
+Version: 2.7.12
 Summary: ValidMind Library
 License: Commercial License
 Author: Andres Rodriguez
@@ -15,8 +15,8 @@ Provides-Extra: all
 Provides-Extra: huggingface
 Provides-Extra: llm
 Provides-Extra: pytorch
-Provides-Extra: r-support
 Requires-Dist: aiohttp[speedups]
+Requires-Dist: anywidget (>=0.9.13,<0.10.0)
 Requires-Dist: arch
 Requires-Dist: bert-score (>=0.3.13)
 Requires-Dist: catboost
@@ -35,14 +35,13 @@ Requires-Dist: numba (<0.59.0)
 Requires-Dist: numpy
 Requires-Dist: openai (>=1)
 Requires-Dist: pandas (>=1.1,<=2.0.3)
-Requires-Dist: plotly
+Requires-Dist: plotly (<6.0.0)
 Requires-Dist: plotly-express
 Requires-Dist: polars
 Requires-Dist: pycocoevalcap (>=1.2,<2.0) ; extra == "all" or extra == "llm"
 Requires-Dist: python-dotenv
 Requires-Dist: ragas (>=0.2.3) ; extra == "all" or extra == "llm"
 Requires-Dist: rouge (>=1)
-Requires-Dist: rpy2 (>=3.5.10,<4.0.0) ; extra == "all" or extra == "r-support"
 Requires-Dist: scikit-learn (<1.6.0)
 Requires-Dist: scipy
 Requires-Dist: scorecardpy (>=0.1.9.6,<0.2.0.0)
@@ -53,6 +52,7 @@ Requires-Dist: shap (==0.44.1)
 Requires-Dist: statsmodels
 Requires-Dist: tabulate (>=0.8.9,<0.9.0)
 Requires-Dist: textblob (>=0.18.0.post0,<0.19.0)
+Requires-Dist: tiktoken
 Requires-Dist: torch (>=1.10.0) ; extra == "all" or extra == "llm" or extra == "pytorch"
 Requires-Dist: tqdm
 Requires-Dist: transformers (>=4.32.0,<5.0.0) ; extra == "all" or extra == "huggingface" or extra == "llm"

{validmind-2.7.10 → validmind-2.7.12}/pyproject.toml

@@ -10,10 +10,11 @@ description = "ValidMind Library"
 license = "Commercial License"
 name = "validmind"
 readme = "README.pypi.md"
-version = "2.7.10"
+version = "2.7.12"
 
 [tool.poetry.dependencies]
 aiohttp = {extras = ["speedups"], version = "*"}
+anywidget = "^0.9.13"
 arch = "*"
 bert-score = ">=0.3.13"
 catboost = "*"
@@ -32,7 +33,7 @@ numba = "<0.59.0" # TODO: https://github.com/validmind/validmind-library/pull/28
 numpy = "*"
 openai = ">=1"
 pandas = ">=1.1,<=2.0.3"
-plotly = "*"
+plotly = "<6.0.0"
 plotly-express = "*"
 polars = "*"
 pycocoevalcap = {version = "^1.2", optional = true}
@@ -40,7 +41,6 @@ python = ">=3.8.1,<3.12"
 python-dotenv = "*"
 ragas = {version = ">=0.2.3", optional = true}
 rouge = ">=1"
-rpy2 = {version = "^3.5.10", optional = true}
 scikit-learn = "*,<1.6.0"
 scipy = "*"
 scorecardpy = "^0.1.9.6"
@@ -51,6 +51,7 @@ shap = "0.44.1"
 statsmodels = "*"
 tabulate = "^0.8.9"
 textblob = "^0.18.0.post0"
+tiktoken = "*"
 torch = {version = ">=1.10.0", optional = true}
 tqdm = "*"
 transformers = {version = "^4.32.0", optional = true}
@@ -76,7 +77,6 @@ twine = "^4.0.2"
 
 [tool.poetry.extras]
 all = [
-  "rpy2",
   "torch",
   "transformers",
   "pycocoevalcap",
@@ -94,7 +94,6 @@ llm = [
   "langchain-openai",
 ]
 pytorch = ["torch"]
-r-support = ["rpy2"]
 
 [build-system]
 build-backend = "poetry.core.masonry.api"

validmind-2.7.12/validmind/__version__.py

@@ -0,0 +1 @@
+__version__ = "2.7.12"

{validmind-2.7.10 → validmind-2.7.12}/validmind/errors.py

@@ -228,7 +228,7 @@ class MissingRExtrasError(BaseError):
     def description(self, *args, **kwargs):
         return (
             self.message
-            or "ValidMind r-support needs to be installed: `pip install validmind[r-support]`"
+            or "`rpy2` is required to use R models. Please install it with `pip install rpy2`"
         )
 
 

{validmind-2.7.10 → validmind-2.7.12}/validmind/utils.py

@@ -110,6 +110,7 @@ class NumpyEncoder(json.JSONEncoder):
             self.is_numpy_ndarray: lambda obj: obj.tolist(),
             self.is_numpy_bool: lambda obj: bool(obj),
             self.is_pandas_timestamp: lambda obj: str(obj),
+            self.is_numpy_datetime64: lambda obj: str(obj),
             self.is_set: lambda obj: list(obj),
             self.is_quantlib_date: lambda obj: obj.ISO(),
             self.is_generic_object: self.handle_generic_object,
@@ -142,6 +143,9 @@ class NumpyEncoder(json.JSONEncoder):
     def is_pandas_timestamp(self, obj):
         return isinstance(obj, pd.Timestamp)
 
+    def is_numpy_datetime64(self, obj):
+        return isinstance(obj, np.datetime64)
+
     def is_set(self, obj):
         return isinstance(obj, set)
 
@@ -152,11 +156,12 @@ class NumpyEncoder(json.JSONEncoder):
         return isinstance(obj, object)
 
     def handle_generic_object(self, obj):
-        return (
-            obj.__str__()
-            if type(obj).__dict__.get("__str__")
-            else str(obj).split(".")[1].split(" ")[0]
-        )
+        try:
+            if hasattr(obj, "__str__"):
+                return obj.__str__()
+            return obj.__class__.__name__
+        except Exception:
+            return str(type(obj).__name__)
 
     def encode(self, obj):
         obj = nan_to_none(obj)
@@ -177,6 +182,18 @@ class HumanReadableEncoder(NumpyEncoder):
             else obj.tolist()
         )
 
+    def default(self, obj):
+        if self.is_dataframe(obj):
+            return {
+                "type": str(type(obj)),
+                "preview": obj.head(5).to_dict(orient="list"),
+                "shape": f"{obj.shape[0]} rows x {obj.shape[1]} columns",
+            }
+        return super().default(obj)
+
+    def is_dataframe(self, obj):
+        return isinstance(obj, pd.DataFrame)
+
 
 def get_full_typename(o: Any) -> Any:
     """We determine types based on type names so we don't have to import

{validmind-2.7.10 → validmind-2.7.12}/validmind/vm_models/result/utils.py

@@ -44,31 +44,12 @@ async def update_metadata(content_id: str, text: str, _json: Union[Dict, List] =
     content_id = parts[0]
     revision_name = parts[1] if len(parts) > 1 else None
 
-    # we always want composite metric definitions to be updated
-    should_update = content_id.startswith("composite_metric_def:")
-
-    # if we are updating a metric or test description, we check if the text
-    # has changed from the last time it was logged, and only update if it has
-    if content_id.split(":", 1)[0] in ["metric_description", "test_description"]:
-        try:
-            md = await api_client.aget_metadata(content_id)
-            # if there is an existing description, only update it if the new one
-            # is different and is an AI-generated description
-            should_update = (
-                md["text"] != text if revision_name == AI_REVISION_NAME else False
-            )
-            logger.debug(f"Check if description has changed: {should_update}")
-        except Exception:
-            # if exception, assume its not created yet TODO: don't catch all
-            should_update = True
-
-    if should_update:
-        if revision_name:
-            content_id = f"{content_id}::{revision_name}"
+    if revision_name:
+        content_id = f"{content_id}::{revision_name}"
 
-        logger.debug(f"Updating metadata for `{content_id}`")
+    logger.debug(f"Updating metadata for `{content_id}`")
 
-        await api_client.alog_metadata(content_id, text, _json)
+    await api_client.alog_metadata(content_id, text, _json)
 
 
 def check_for_sensitive_data(data: pd.DataFrame, inputs: List[VMInput]):

validmind-2.7.10/validmind/__version__.py

@@ -1 +0,0 @@
-__version__ = "2.7.10"