validmind 2.4.4__tar.gz → 2.4.6__tar.gz

{validmind-2.4.4 → validmind-2.4.6}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: validmind
-Version: 2.4.4
+Version: 2.4.6
 Summary: ValidMind Developer Framework
 License: Commercial License
 Author: Andres Rodriguez
@@ -32,7 +32,7 @@ Requires-Dist: mistune (>=3.0.2,<4.0.0)
 Requires-Dist: nltk (>=3.8.1,<4.0.0)
 Requires-Dist: numba (<0.59.0)
 Requires-Dist: numpy
-Requires-Dist: openai (>=1) ; extra == "all" or extra == "llm"
+Requires-Dist: openai (>=1) ; extra == "all"
 Requires-Dist: pandas (>=1.1,<2)
 Requires-Dist: plotly
 Requires-Dist: plotly-express

{validmind-2.4.4 → validmind-2.4.6}/pyproject.toml

@@ -10,7 +10,7 @@ description = "ValidMind Developer Framework"
 license = "Commercial License"
 name = "validmind"
 readme = "README.pypi.md"
-version = "2.4.4"
+version = "2.4.6"
 
 [tool.poetry.dependencies]
 aiohttp = {extras = ["speedups"], version = "*"}
@@ -29,8 +29,8 @@ mistune = "^3.0.2"
 nltk = "^3.8.1"
 numba = "<0.59.0" # TODO: https://github.com/validmind/developer-framework/pull/28
 numpy = "*"
-openai = {version = ">=1", optional = true}
-pandas = ">=1.1,<2" # TODO small issues blocking 2.0 but it would speed things up to use arrow backend
+openai = ">=1"
+pandas = ">=1.1,<2" # TODO: small issues blocking 2.0 but it would speed things up to use arrow backend
 plotly = "*"
 plotly-express = "*"
 polars = "*"
@@ -64,7 +64,6 @@ flake8 = "^4.0.1"
 ipykernel = "^6.22.0"
 isort = "^5.12.0"
 jupyter = "^1.0.0"
-openai = "^1.3.7"
 papermill = "^2.4.0"
 pdoc = "^14.4.0"
 pre-commit = "^3.3.3"
@@ -88,7 +87,6 @@ huggingface = ["transformers", "sentencepiece"]
 llm = [
   "torch",
   "transformers",
-  "openai",
   "pycocoevalcap",
   "ragas",
   "sentencepiece",

validmind-2.4.6/validmind/__version__.py

@@ -0,0 +1 @@
+__version__ = "2.4.6"

{validmind-2.4.4 → validmind-2.4.6}/validmind/tests/data_validation/IQROutliersBarPlot.py

@@ -4,7 +4,6 @@
 
 from dataclasses import dataclass
 
-import numpy as np
 import plotly.graph_objects as go
 
 from validmind.vm_models import Figure, Metric
@@ -62,22 +61,27 @@ class IQROutliersBarPlot(Metric):
 
     name = "iqr_outliers_bar_plot"
     required_inputs = ["dataset"]
-    default_params = {"threshold": 1.5, "num_features": None, "fig_width": 800}
+    default_params = {"threshold": 1.5, "fig_width": 800}
     tasks = ["classification", "regression"]
     tags = ["tabular_data", "visualization", "numerical_data"]
 
     def run(self):
         df = self.inputs.dataset.df
-        num_features = self.params["num_features"]
+
+        # Select numerical features
+        features = self.inputs.dataset.feature_columns_numeric
+
+        # Select non-binary features
+        features = [
+            feature
+            for feature in features
+            if len(self.inputs.dataset.df[feature].unique()) > 2
+        ]
+
         threshold = self.params["threshold"]
         fig_width = self.params["fig_width"]
 
-        # If num_features is None, use all numeric columns.
-        # Otherwise, only use the columns provided in num_features.
-        if num_features is None:
-            df = df.select_dtypes(include=[np.number])
-        else:
-            df = df[num_features]
+        df = df[features]
 
         return self.detect_and_visualize_outliers(df, threshold, fig_width)
 
@@ -98,6 +102,9 @@ class IQROutliersBarPlot(Metric):
             # Compute outliers
             outliers = self.compute_outliers(df[col], threshold)
 
+            if outliers.empty:
+                continue  # Skip plotting if there are no outliers
+
             Q1_count = outliers[
                 (outliers >= 0) & (outliers < outliers.quantile(0.25))
             ].count()

{validmind-2.4.4 → validmind-2.4.6}/validmind/tests/data_validation/IQROutliersTable.py

@@ -4,7 +4,6 @@
 
 from dataclasses import dataclass
 
-import numpy as np
 import pandas as pd
 
 from validmind.vm_models import Metric, ResultSummary, ResultTable, ResultTableMetadata
@@ -53,12 +52,22 @@ class IQROutliersTable(Metric):
 
     name = "iqr_outliers_table"
     required_inputs = ["dataset"]
-    default_params = {"features": None, "threshold": 1.5}
+    default_params = {"threshold": 1.5}
     tasks = ["classification", "regression"]
     tags = ["tabular_data", "numerical_data"]
 
     def run(self):
-        features = self.params["features"]
+
+        # Select numerical features
+        features = self.inputs.dataset.feature_columns_numeric
+
+        # Select non-binary features
+        features = [
+            feature
+            for feature in features
+            if len(self.inputs.dataset.df[feature].unique()) > 2
+        ]
+
         threshold = self.params["threshold"]
 
         df = self.inputs.dataset.df
@@ -80,9 +89,7 @@ class IQROutliersTable(Metric):
         upper_bound = Q3 + threshold * IQR
         return series[(series < lower_bound) | (series > upper_bound)]
 
-    def detect_and_analyze_outliers(self, df, features=None, threshold=1.5):
-        if features is None:
-            features = df.select_dtypes(include=[np.number]).columns.tolist()
+    def detect_and_analyze_outliers(self, df, features, threshold=1.5):
 
         outliers_summary = []
         for feature in features:

{validmind-2.4.4 → validmind-2.4.6}/validmind/tests/run.py

@@ -89,22 +89,24 @@ def _update_plotly_titles(figures, input_groups, title_template):
     for i, figure in enumerate(figures):
         figure.figure.layout.title.text = title_template.format(
             current_title=f"{current_title} " if current_title else "",
-            input_description=", ".join(
-                f"{k}={v if isinstance(v, str) else v.input_id}"
+            input_description=" and ".join(
+                f"{k}: {v if isinstance(v, str) else ', '.join(item.input_id for item in v) if isinstance(v, list) and all(hasattr(item, 'input_id') for item in v) else v.input_id}"
                 for k, v in input_groups[i].items()
             ),
         )
 
 
 def _update_matplotlib_titles(figures, input_groups, title_template):
-    current_title = figures[0].figure.get_title()
+    current_title = (
+        figures[0].figure._suptitle.get_text() if figures[0].figure._suptitle else ""
+    )
 
     for i, figure in enumerate(figures):
         figure.figure.suptitle(
             title_template.format(
                 current_title=f"{current_title} " if current_title else "",
                 input_description=" and ".join(
-                    f"{k}: {v if isinstance(v, str) else v.input_id}"
+                    f"{k}: {v if isinstance(v, str) else ', '.join(item.input_id for item in v) if isinstance(v, list) and all(hasattr(item, 'input_id') for item in v) else v.input_id}"
                     for k, v in input_groups[i].items()
                 ),
             )
@@ -139,10 +141,20 @@ def metric_comparison(
     """Build a comparison result for multiple metric results"""
     ref_id = str(uuid4())
 
-    input_group_strings = [
-        {k: v if isinstance(v, str) else v.input_id for k, v in group.items()}
-        for group in input_groups
-    ]
+    input_group_strings = []
+
+    for group in input_groups:
+        new_group = {}
+        for k, v in group.items():
+            if isinstance(v, str):
+                new_group[k] = v
+            elif hasattr(v, "input_id"):
+                new_group[k] = v.input_id
+            elif isinstance(v, list) and all(hasattr(item, "input_id") for item in v):
+                new_group[k] = ", ".join([item.input_id for item in v])
+            else:
+                raise ValueError(f"Unsupported type for value: {v}")
+        input_group_strings.append(new_group)
 
     merged_summary = _combine_summaries(
         [
@@ -173,9 +185,11 @@ def metric_comparison(
             ),
         ],
         inputs=[
-            input if isinstance(input, str) else input.input_id
+            item.input_id if hasattr(item, "input_id") else item
             for group in input_groups
             for input in group.values()
+            for item in (input if isinstance(input, list) else [input])
+            if hasattr(item, "input_id") or isinstance(item, str)
         ],
         output_template=output_template,
         metric=MetricResult(
@@ -198,10 +212,20 @@ def threshold_test_comparison(
     """Build a comparison result for multiple threshold test results"""
     ref_id = str(uuid4())
 
-    input_group_strings = [
-        {k: v if isinstance(v, str) else v.input_id for k, v in group.items()}
-        for group in input_groups
-    ]
+    input_group_strings = []
+
+    for group in input_groups:
+        new_group = {}
+        for k, v in group.items():
+            if isinstance(v, str):
+                new_group[k] = v
+            elif hasattr(v, "input_id"):
+                new_group[k] = v.input_id
+            elif isinstance(v, list) and all(hasattr(item, "input_id") for item in v):
+                new_group[k] = ", ".join([item.input_id for item in v])
+            else:
+                raise ValueError(f"Unsupported type for value: {v}")
+        input_group_strings.append(new_group)
 
     merged_summary = _combine_summaries(
         [

validmind-2.4.4/validmind/__version__.py

@@ -1 +0,0 @@
-__version__ = "2.4.4"