Update task type logic for dataframes / datasets

Author: whoseoyster

examples/tabular-classification/churn-classifier/churn-classifier-sklearn.ipynb

@@ -254,7 +254,6 @@
     "    class_names=class_names,\n",
     "    label_column_name='churn',\n",
     "    commit_message='first version',\n",
-    "    task_type=TaskType.TabularClassification,\n",
     "    feature_names=feature_names,\n",
     "    categorical_feature_names=[\"Gender\", \"Geography\"],\n",
     ")\n",

unboxapi/__init__.py

@@ -447,6 +447,7 @@ def add_model(
                     endpoint = f"projects/{project_id}/ml-models"
                     payload = dict(
                         name=name,
+                        taskType=task_type.value,
                         commitMessage=commit_message,
                         classNames=class_names,
                         architectureType=model_type.name,

unboxapi/projects.py

@@ -51,9 +51,11 @@ def add_dataset(
         *args,
         **kwargs,
     ) -> Dataset:
-        kwargs["project_id"] = self.id
-        return self.client.add_dataset(*args, **kwargs)
+        return self.client.add_dataset(
+            *args, project_id=self.id, task_type=tasks.TaskType(self.taskType), **kwargs
+        )
 
     def add_dataframe(self, *args, **kwargs) -> Dataset:
-        kwargs["project_id"] = self.id
-        return self.client.add_dataframe(*args, **kwargs)
+        return self.client.add_dataframe(
+            *args, project_id=self.id, task_type=tasks.TaskType(self.taskType), **kwargs
+        )