validmind 2.4.10__py3-none-any.whl → 2.5.1__py3-none-any.whl

validmind/__version__.py

@@ -1 +1 @@
-__version__ = "2.4.10"
+__version__ = "2.5.1"

validmind/api_client.py

@@ -152,6 +152,7 @@ def _get_session() -> aiohttp.ClientSession:
                 "X-API-KEY": _api_key,
                 "X-API-SECRET": _api_secret,
                 "X-PROJECT-CUID": _model_cuid,
+                "X-MONITORING": str(_monitoring),
             }
         )
 

validmind/client.py

@@ -48,7 +48,6 @@ def init_dataset(
     index_name: str = None,
     date_time_index: bool = False,
     columns: list = None,
-    options: dict = None,
     text_column: str = None,
     target_column: str = None,
     feature_columns: list = None,
@@ -72,7 +71,6 @@ def init_dataset(
     Args:
         dataset : dataset from various python libraries
         model (VMModel): ValidMind model object
-        options (dict): A dictionary of options for the dataset
         targets (vm.vm.DatasetTargets): A list of target variables
         target_column (str): The name of the target column in the dataset
         feature_columns (list): A list of names of feature columns in the dataset

validmind/input_registry.py

@@ -6,6 +6,8 @@
 Central class to register inputs
 """
 
+from validmind.vm_models.input import VMInput
+
 from .errors import InvalidInputError
 
 
@@ -14,6 +16,12 @@ class InputRegistry:
         self.registry = {}
 
     def add(self, key, obj):
+        if not isinstance(obj, VMInput):
+            raise InvalidInputError(
+                f"Input object must be an instance of VMInput. "
+                f"Got {type(obj)} instead."
+            )
+
         self.registry[key] = obj
 
     def get(self, key):

validmind/tests/__types__.py

@@ -113,6 +113,10 @@ TestID = Literal[
     "validmind.model_validation.statsmodels.GINITable",
     "validmind.model_validation.statsmodels.RegressionModelForecastPlot",
     "validmind.model_validation.statsmodels.DurbinWatsonTest",
+    "validmind.ongoing_monitoring.PredictionCorrelation",
+    "validmind.ongoing_monitoring.PredictionAcrossEachFeature",
+    "validmind.ongoing_monitoring.FeatureDrift",
+    "validmind.ongoing_monitoring.TargetPredictionDistributionPlot",
     "validmind.data_validation.MissingValuesRisk",
     "validmind.data_validation.IQROutliersTable",
     "validmind.data_validation.BivariateFeaturesBarPlots",

validmind/tests/data_validation/DatasetDescription.py

@@ -134,6 +134,7 @@ class DatasetDescription(Metric):
                     )
             else:
                 vm_dataset_variables[column] = {"id": column, "type": str(type)}
+
         return list(vm_dataset_variables.values())
 
     def describe_dataset_field(self, df, field):

validmind/tests/model_validation/sklearn/ClassifierPerformance.py

@@ -4,7 +4,7 @@
 
 from dataclasses import dataclass
 
-from numpy import unique
+import numpy as np
 from sklearn.metrics import classification_report, roc_auc_score
 from sklearn.preprocessing import LabelBinarizer
 
@@ -71,7 +71,7 @@ class ClassifierPerformance(Metric):
         When building a multi-class summary we need to calculate weighted average,
         macro average and per class metrics.
         """
-        classes = {str(i) for i in unique(self.inputs.dataset.y)}
+        classes = {str(i) for i in np.unique(self.inputs.dataset.y)}
         pr_f1_table = [
             {
                 "Class": class_name,
@@ -126,9 +126,18 @@ class ClassifierPerformance(Metric):
             output_dict=True,
             zero_division=0,
         )
-        report["roc_auc"] = multiclass_roc_auc_score(
-            self.inputs.dataset.y,
-            self.inputs.dataset.y_pred(self.inputs.model),
-        )
+
+        y_true = self.inputs.dataset.y
+
+        if len(np.unique(y_true)) > 2:
+            y_pred = self.inputs.dataset.y_pred(self.inputs.model)
+            y_true = y_true.astype(y_pred.dtype)
+            roc_auc = self.multiclass_roc_auc_score(y_true, y_pred)
+        else:
+            y_prob = self.inputs.dataset.y_prob(self.inputs.model)
+            y_true = y_true.astype(y_prob.dtype).flatten()
+            roc_auc = roc_auc_score(y_true, y_prob)
+
+        report["roc_auc"] = roc_auc
 
         return self.cache_results(report)

validmind/tests/model_validation/sklearn/ClusterPerformance.py

@@ -57,7 +57,7 @@ class ClusterPerformance(Metric):
         "model_performance",
     ]
 
-    def cluser_performance_metrics(
+    def cluster_performance_metrics(
         self, y_true_train, y_pred_train, y_true_test, y_pred_test, samples, metric_info
     ):
         y_true_train = y_true_train.astype(y_pred_train.dtype).flatten()
@@ -107,7 +107,7 @@ class ClusterPerformance(Metric):
         y_true_test = y_true_test.astype(class_pred_test.dtype)
 
         samples = ["train", "test"]
-        results = self.cluser_performance_metrics(
+        results = self.cluster_performance_metrics(
             y_true_train,
             class_pred_train,
             y_true_test,

validmind/tests/model_validation/sklearn/MinimumROCAUCScore.py

@@ -5,6 +5,7 @@
 from dataclasses import dataclass
 from typing import List
 
+import numpy as np
 import pandas as pd
 from sklearn import metrics, preprocessing
 
@@ -99,9 +100,15 @@ class MinimumROCAUCScore(ThresholdTest):
 
     def run(self):
         y_true = self.inputs.dataset.y
-        class_pred = self.inputs.dataset.y_pred(self.inputs.model)
-        y_true = y_true.astype(class_pred.dtype)
-        roc_auc = self.multiclass_roc_auc_score(y_true, class_pred)
+
+        if len(np.unique(y_true)) > 2:
+            class_pred = self.inputs.dataset.y_pred(self.inputs.model)
+            y_true = y_true.astype(class_pred.dtype)
+            roc_auc = self.multiclass_roc_auc_score(y_true, class_pred)
+        else:
+            y_prob = self.inputs.dataset.y_prob(self.inputs.model)
+            y_true = y_true.astype(y_prob.dtype).flatten()
+            roc_auc = metrics.roc_auc_score(y_true, y_prob)
 
         passed = roc_auc > self.params["min_threshold"]
         results = [