added step

HeidiSteen · HeidiSteen · commit 6051552f736b · 2018-11-05T14:27:58.000-08:00
diff --git a/docs/advanced-analytics/tutorials/media/sqldev-python-newcolumn.png b/docs/advanced-analytics/tutorials/media/sqldev-python-newcolumn.png
diff --git a/docs/advanced-analytics/tutorials/sqldev-in-database-r-for-sql-developers.md b/docs/advanced-analytics/tutorials/sqldev-in-database-r-for-sql-developers.md
@@ -13,9 +13,10 @@ manager: cgronlun
 # Tutorial: In-Database R analytics for SQL developers
 [!INCLUDE[appliesto-ss-xxxx-xxxx-xxx-md-winonly](../../includes/appliesto-ss-xxxx-xxxx-xxx-md-winonly.md)]
 
-In this tutorial for SQL programmers, you gain hands-on experience using the R language to build and deploy a machine learning solution by wrapping R code in stored procedures.
+In this tutorial for SQL programmers, learn about R integration by building and deploying an R-based machine learning solution using a [NYCTaxi_sample](demo-data-nyctaxi-in-sql.md) database on SQL Server. 
+
+This tutorial introduces you to R functions used in a data modeling workflow. Steps include data exploration, building and training a binary classification model, and model deployment. You'll use sample data from the New York City Taxi and Limosine Commission, and the model you will build predicts whether a trip is likely to result in a tip based on the time of day, distance travelled, and pick-up location. All of the R code used in this tutorial is wrapped in stored procedures that you create and run in Management Studio.
 
-Using sample data from the New York City Taxi and Limosine Commission, you will build a binary classification model that predicts whether a particular trip is likely to get a tip or not, based on columns such as the time of day, distance, and pick-up location.
 
 > [!NOTE]
 > 
diff --git a/docs/advanced-analytics/tutorials/sqldev-py5-train-and-save-a-model-using-t-sql.md b/docs/advanced-analytics/tutorials/sqldev-py5-train-and-save-a-model-using-t-sql.md
@@ -48,6 +48,16 @@ You load the modules and call the necessary functions to create and train the mo
     GO
     ```
 
+## Add a name column in nyc_taxi_models
+
+Scripts in this tutorial store a model name as a label for generated models. The model name is used in queries to select a revoscalepy or SciKit model.
+
+1. In Management Studio, open the **nyc_taxi_models** table.
+
+2. Right-click **Columns** and click **New Column**. Set the column name to *name*, with a type **nchar(250)**, and allow nulls.
+
+    ![Name column for storing model names](media/sqldev-python-newcolumn.png)
+
 ## Build a logistic regression model
 
 After the data has been prepared, you can use it to train a model. You do this by calling a stored procedure that runs some Python code, taking as input the training data table. For this tutorial, you create two models, both binary classification models:
diff --git a/docs/advanced-analytics/tutorials/sqldev-py6-operationalize-the-model.md b/docs/advanced-analytics/tutorials/sqldev-py6-operationalize-the-model.md
@@ -59,7 +59,6 @@ import numpy;
 from sklearn import metrics
 
 mod = pickle.loads(lmodel2)
-
 X = InputDataSet[["passenger_count", "trip_distance", "trip_time_in_secs", "direct_distance"]]
 y = numpy.ravel(InputDataSet[["tipped"]])
 
@@ -109,12 +108,13 @@ X = InputDataSet[["passenger_count", "trip_distance", "trip_time_in_secs", "dire
 y = numpy.ravel(InputDataSet[["tipped"]])
 
 probArray = rx_predict(mod, X)
-prob_list = prob_array["tipped_Pred"].values 
+probList = probArray["tipped_Pred"].values 
 
 probArray = numpy.asarray(probList)
 fpr, tpr, thresholds = metrics.roc_curve(y, probArray)
 aucResult = metrics.auc(fpr, tpr)
 print ("AUC on testing data is: " + str(aucResult))
+
 OutputDataSet = pandas.DataFrame(data = probList, columns = ["predictions"])
 ',	
   @input_data_1 = @inquery,
