ray-project
diff --git a/‎doc/source/ray-overview/examples/e2e-timeseries/e2e_timeseries/02-Validation.ipynb‎
Lines changed: 13 additions & 2 deletions b/‎doc/source/ray-overview/examples/e2e-timeseries/e2e_timeseries/02-Validation.ipynb‎
Lines changed: 13 additions & 2 deletions
diff --git a/‎doc/source/ray-overview/examples/e2e-timeseries/e2e_timeseries/__init__.py‎ b/‎doc/source/ray-overview/examples/e2e-timeseries/e2e_timeseries/__init__.py‎
@@ -52,6 +52,19 @@
     "**Note**: Ray Data operates as batch processing with streaming execution rather than a real-time stream processing engine like Flink or Kafka Streams. This approach proves especially useful for iterative ML workloads, ETL pipelines, and preprocessing before training or inference. Ray typically delivers a [**2-17x throughput improvement**](https://www.anyscale.com/blog/offline-batch-inference-comparing-ray-apache-spark-and-sagemaker#-results-of-throughput-from-experiments) over solutions like Spark and SageMaker Batch Transform.\n"
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Enable importing from e2e_timeseries module.\n",
+    "import os\n",
+    "import sys\n",
+    "\n",
+    "sys.path.insert(0, os.path.abspath(os.path.join(os.getcwd(), os.pardir)))"
+   ]
+  },
   {
    "cell_type": "markdown",
    "metadata": {},
@@ -65,8 +78,6 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "import os\n",
-    "\n",
     "import numpy as np\n",
     "import ray\n",
     "import torch\n",