Use testData instead of trainingData for inference. (aws#450)

nadiaya · web-flow · commit 13abcfbe7b31 · 2018-10-31T09:43:25.000-07:00
diff --git a/sagemaker-spark/pyspark_mnist/pyspark_mnist_xgboost.ipynb b/sagemaker-spark/pyspark_mnist/pyspark_mnist_xgboost.ipynb
@@ -172,7 +172,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "transformedData = model.transform(trainingData)\n",
+    "transformedData = model.transform(testData)\n",
     "\n",
     "transformedData.show()"
    ]

Original file line number	Diff line number	Diff line change
`@@ -172,7 +172,7 @@`
`172`	`172`	`"metadata": {},`
`173`	`173`	`"outputs": [],`
`174`	`174`	`"source": [`
`175`		`- "transformedData = model.transform(trainingData)\n",`
	`175`	`+ "transformedData = model.transform(testData)\n",`
`176`	`176`	`"\n",`
`177`	`177`	`"transformedData.show()"`
`178`	`178`	`]`