Skip to Content
Technical Articles

SAP Data Intelligence: Create your first ML Scenario

You may know that SAP Data Intelligence is SAP’s new platform to create and deploy Machine Learning (ML) projects. Whether your data is structured or unstructured, in SAP HANA or outside, SAP Data Intelligence is your new friend.

This blog shows how to create a very basic ML project in SAP Data Intelligence, We will train a linear regression model in Python with just a single predictor column. Based on how many minutes it takes a person run a half-marathon, the model will estimate the person’s marathon time. That trained model will then be exposed as REST-API for inference. Your business applications could now leverage these predictions in your day-to-day processes. After this introductory tutorial you will be familiar with the core components of an ML scenario in SAP Data Intelligence, and the implementation of more advanced requirements is likely to follow a similar pattern.

For those who prefer R instead of Python, Ingo Peter has published a blog which implements this  first ML Scenario with R.

Should you have access to a system you can follow hands-on yourself. You can currently request a trial or talk to your Account Executive anytime.

Updated on 26 April 2020 for SAP Data Intelligence version 3

Table of contents


Training data

Our dataset is a small CSV file, which contains the running times (in minutes) of 117 people, who ran both the Zurich Marathon as well as a local Half-Marathon in the same region, the Greifenseelauf. The first rows show the fastest runners:

1 73 149
2 74 154
3 78 158
4 73 165
5 74 172


SAP Data Intelligence can connect to range of sources. However, to follow this tutorial hands-on, please place this this file into your own Amazon S3 bucket.

Alternatively, you can also store the file in SAP Data Intelligence’s internal data lake. Just follow the advice of Philipp Zaltenbach in this comment on which adjustments are needed should you prefer that option. This comment explains how to load the data from SAP HANA into the training pipeline.

In case you have placed the file into an S3 bucket, you need to know:

  • the endpoint URL of the S3 server, without your bucket name, ie
  • the authentication region, ie us-east-1
  • the access key and secret key

Data connection

Logon to SAP Data Intelligence and you will see the different application tiles. These tiles can be rearranged, hence in your system the sequence of the tiles might be different. Click into “Connection Management”, where you can centrally define a connection to the S3 bucket that holds the data.


You may see a number of connections that have already been set up. Click the “+” sign on top to create a new connection. Set the “Connection Type” to “S3” and set the id to “s3_files”.


and set the “Connection Type” to “S3”. Enter the bucket’s endpoint, the region and the access key as well as the secret key. Also, set the “Root Path” to the name of your bucket.

Verify the settings with “Test Connection” and click “Create”.


Data exploration and free-style Data Science

Now you are ready to work with the data in SAP Data Intelligence. Go back to the main page of SAP Data Intelligence and click into the “ML Scenario Manager”, where you will carry out all Machine Learning related activities.

Click the little “+”-sign on the top right to create a new scenario. Name it “Marathon time predictions” and you enter further details into the “Business Question” section. Click “Create”.


You see the empty scenario. You will use the Notebooks to explore the data and to script the regression model in Python. Pipelines bring the code into production. Executions of these pipelines will create Machine Learning models, which are then deployed as REST-API for inference.


But one step at a time. Next, you load and explore the data in a Notebook. Select the “Notebooks” tab, then click the “+”-sign.  Name the Notebook “10 Data exploration and model training”. Click “Create” and the Notebook opens up. You will be prompted for the kernel, keep the default of “Python 3”.


Now you are free to script in Python, to explore the data and train the regression model. The central connection to your S3 bucket can be leveraged by accessing the keys.

import notebook_hana_connector.notebook_hana_connector
di_connection = notebook_hana_connector.notebook_hana_connector.get_datahub_connection(id_="s3_files")
access_key = di_connection["contentData"]["accessKey"]
secret_key = di_connection["contentData"]["secretKey"]


Install the library that is required for Python to connect to the bucket.

!pip install boto3


The CSV file can then be loaded into a pandas DataFrame. Just make sure that the value assigned to the bucket corresponds with your own bucket name.

import boto3
import pandas as pd
import io
client = boto3.client('s3', aws_access_key_id=access_key, 
bucket = 'i056450-datasets'
object_key = 'RunningTimes.csv'
csv_obj = client.get_object(Bucket=bucket, Key=object_key)
body = csv_obj['Body']
csv_string ='utf-8')
df_data = pd.read_csv(io.StringIO(csv_string), sep=";")


Use Python to explore the data. Display the first few rows for example.



Or plot the runners’ half-marathon time against the full marathon time.

y_true = df_data["MARATHON_MINUTES"]

%matplotlib inline
import matplotlib.pyplot as plot
plot.scatter(x, y_true, color = 'darkblue');
plot.xlabel("Minutes Half-Marathon");
plot.ylabel("Minutes Marathon");


There is clearly a linear relationship. No surprise, If you can run a marathon fast, you are also likely to be one of the faster half-marathon runners.


Train regression model

Begin by installing the scikit-learn library, which is very popular for Machine Learning in Python on tabular data such as ours.

!pip install sklearn​


Then continue by training the linear regression model with the newly installed scikit-learn library. In case you are not familiar with a linear regression, you might enjoy the openSAP course “Introduction to Statistics for Data Science“.

from sklearn.linear_model import LinearRegression
lm = LinearRegression(), y_true)


Plot the linear relationship that was estimated.

plot.scatter(x, y_true, color = 'darkblue');
plot.plot(x, lm.predict(x), color = 'red');
plot.xlabel("Actual Minutes Half-Marathon");
plot.ylabel("Actual Minutes Marathon");


Calculate the Root Mean Squared Error (RMSE) as quality indicator of the model’s performance on the training data. This indicator will be shown later on in the ML Scenario Manager.

In this introductory example the RMSE is calculated rather manually. This way, I hope, it easier to understand its meaning, in case you have not yet been familiar with it. Most Data Scientists would probably leverage a Python package to shorten this bit.

import numpy as np
y_pred = lm.predict(x)
mse = np.mean((y_pred - y_true)**2)
rmse = np.sqrt(mse)
rmse = round(rmse, 2)
print("RMSE: " , str(rmse))
print("n: ", str(len(x)))


A statistician may want to run further tests on the regression, ie on the distribution of the errors. We skip these steps in this example.


Save regression model

You could apply the model immediately on a new observation. However, to deploy the model later on, we will be using one graphical pipeline to train and save the model and a second pipeline for inference.

Hence at this point we also spread the Python code across two separate Notebooks for clarity.  Therefore we finish this Notebook by saving the model. The production pipeline will save the model as pickled object. Hence the model is also saved here in the Notebook as pickle object as well.

import pickle
pickle.dump(lm, open("marathon_lm.pickle.dat", "wb"))


Don’t forget to save the Notebook itself as well.


Load and apply regression model

To create the second Notebook, go back to the overview page of your ML Scenario and click the “+”-sign in the Notebooks section.


Name the new Notebook “20 Apply model”, confirm “Python 3” as kernel and the empty Notebook opens up. Use the following code to load the model that has just been saved.

import pickle
lm_loaded = pickle.load(open("marathon_lm.pickle.dat", "rb"))


It’s time to apply the model. Predict a runner’s marathon time if the person runs a half-marathon in 2 hours.

x_new = 120
predictions = lm_loaded.predict([[x_new]])
round(predictions[0], 2)


The model estimates a time of just under 4 hours and 24 minutes.



Now everything is in place to start deploying the model in two graphical pipelines.

  • One pipeline to train the model and save it into the ML Scenario.
  • And Another pipeline to surface the model as REST-API for inference


Training pipeline

To create the graphical pipeline to retrain the model, go to your ML Scenario’s main page, select the “Pipelines” tab and click the “+”-sign.


Name the pipeline “10 Train” and select the “Python Producer”-template.


You should see the following pipeline, which we just need to adjust. In principle, the pipeline loads data with the “Read File”-operator. That data is passed to a Python-operator, in which the ML model is trained. The same Python-operator stores the model in the ML Scenario through the “Artifact Producer”. The Python-operator’s second output can pass a quality metric of the model to the same ML Scenario. Once both model and metric are saved, the pipeline’s execution is ended with the “Graph Terminator”.


Now adjust the template to our scenario. Begin with the data connection. Select the “Read File”-operator and click the “Open Configuration” option.


In the Configuration panel on the right-hand side open the “Connection”-configuration, set “Configuration type” to “Connection Management” and you can set the “Connection ID” to the connection you had created earlier in the Connection Management. In this example the connection is called “s3_files”.


Save this configuration. Now open the “Path” setting to specify the name of the file in the S3-bucket.


The files that exist in the S3-bucket are listed. Select the file RunningTimes.csv and click “OK”. Your settings should now be similar to:


Next we adjust the Python code that trains the regression model. Select the “Python 3”-operator and click the Script-option.


The template code opens up. It shows how to pass the model and metrics into the ML Scenario. Replace the whole code with the following. That code receives the data from the “Read File”-operator, uses the code from the Notebook to train the model and passes the trained model as well as its quality indicator (RMSE) to the ML Scenario.

# Example Python script to perform training on input data & generate Metrics & Model Blob
def on_input(data):
    # Obtain data
    import pandas as pd
    import io
    df_data = pd.read_csv(io.StringIO(data), sep=";")
    # Get predictor and target
    x = df_data[["HALFMARATHON_MINUTES"]]
    y_true = df_data["MARATHON_MINUTES"]
    # Train regression
    from sklearn.linear_model import LinearRegression
    lm = LinearRegression(), y_true)
    # Model quality
    import numpy as np
    y_pred = lm.predict(x)
    mse = np.mean((y_pred - y_true)**2)
    rmse = np.sqrt(mse)
    rmse = round(rmse, 2)
    # to send metrics to the Submit Metrics operator, create a Python dictionary of key-value pairs
    metrics_dict = {"RMSE": str(rmse), "n": str(len(df_data))}
    # send the metrics to the output port - Submit Metrics operator will use this to persist the metrics 
    api.send("metrics", api.Message(metrics_dict))

    # create & send the model blob to the output port - Artifact Producer operator will use this to persist the model and create an artifact ID
    import pickle
    model_blob = pickle.dumps(lm)
    api.send("modelBlob", model_blob)
api.set_port_callback("input", on_input)


Close the Script-window, then hit “Save” in the menu bar.


Before running the pipeline, we just need to create a Docker image for the Python operator. This gives the flexibility to leverage virtually any Python library, you just need to provide the Docker file, which installs the necessary libraries. You find the Dockerfiles by clicking into the “Repository”-tab on the left, then right-click the “Dockerfiles” folder and select “Create Docker File”.


Name the file python36marathon.


Enter this code into the Docker File window. This code leverages a base image that comes with SAP Data Intelligence and installs the necessary libraries on it. It’s advised to specify the versions of the libraries to ensure that new versions of these libraries do not impact your environment.

RUN pip3.6 install --user numpy==1.16.4
RUN pip3.6 install --user pandas==0.24.0
RUN pip3.6 install --user sklearn


Update 14 October 2020: A bug slipped into the current release of SAP Data Intelligence cloud, which prevents the build of the Docker file. A fix is on the way, in the meantime please see this temporary workaround (right click on the folder of your Docker File and create the JSON file with “Create File”). Once the file is created and saved, you might need to refresh the browser for the tag to appear in the drop down list of the group.

SAP Data Intelligence uses tags to indicate the content of the Docker File. These tags are used in a graphical pipeline to specify in which Docker image an Operator should run. Open the Configuration panel for the Docker File with the icon on the top-right hand corner.


Add a custom tag to be able to specify that this Docker File is for the Marathon case. Name it  “marathontimes”. No further tags need to be added, as the base image already contains all other tags that are required (“sles”, “python36”, “tornado”).

The “Configuration” should look like:


Now save the Docker file and click the “Build”-icon to start building the Docker image.


Wait a few minutes and you should receive a confirmation that the build completed successfully.


Now you just need to configure the Python operator, which trains the model, to use this Docker image. Go back to the graphical pipeline “10 Train”. Right-click the “Python 3”-operator and select “Group”.


Such a group can contain one or more graphical operators. And on this group level you can specify which Docker image should be used. Select the group, which surrounds the “Python 3” Operator. Now in the group’s Configuration select the tag “marathontimes”. Save the graph.


The pipeline is now complete and we can run it. Go back to the ML Scenario. On the top left you notice a little orange icon, which indicates that the scenario was changed after the current version was created.


Earlier versions of SAP Data Intelligence were only able to execute pipelines that had no changes made to them after the last version was taken. This requirement has been removed, hence there is no need to create a new version at the moment. However, when implementing a productive solution, I strongly suggest to deploy only versioned content.

You can immediately execute this graph, to train and save the model! Select the pipeline in the ML Scenario and click the “Execute” button on the right.


Skip the optional steps until you get to the “Pipeline Parameters”. Set “newArtifactName” to lm_model. The trained regression model will be saved under this name.


Click “Save”. Wait a few seconds until the pipeline executes and completes. Just refresh the page once in a while and you should see the following screen. The metrics section shows the trained model’s quality indicator (RMSE = 16.96) as well as the number of records that were used to train the model (n = 116). The model itself was saved successfully under the name “lm_model”.


If you scroll down on the page, you see how the model’s metrics as well as the model itself have become part of the ML scenario.


We have our model, now we want to use it for real-time inference.

Prediction / Inference with REST-API

Go back to the main page of your ML Scenario and create a second pipeline. This pipeline will provide the REST-API to obtain predictions in real-time. Hence call the pipeline “20 Apply REST-API”. Now select the template “Python Consumer”. This template contains a pipeline that provides a REST-API. We just need to configure it for our model.


The “OpenAPI Servlow” operator provides the REST-API. The “Artifact Consumer” loads the trained model from our ML scenario and the “Python36 – Inference” operator ties the two operators together. It receives the input from the REST-API call (here the half-marathon time) and uses the loaded model to create the prediction, which is then returned by the “OpenAPI Servlow” to the client, which had called the REST-API.


We only need to change the “Python36 – Inference”-operator. Open its “Script” window. At the top of the code, in the on_model() function, replace the single line “model = model_blob” with

import pickle
model = pickle.loads(model_blob)


Add a variable to store the prediction. Look for the on_input() function and add the line “prediction = None” as shown:

def on_input(msg):
    error_message = ""
    success = False
    prediction = None # This line needs to be added


Below the existing comment “# obtain your results” add the syntax to extract the input data (half-marathon time) and to carry out the prediction:

# obtain your results
hm_minutes = json.loads(user_data)['half_marathon_minutes']
prediction = model.predict([[hm_minutes]])


Further at the bottom of the page, change the “if success:” section to:

if success:
    # apply carried out successfully, send a response to the user
    msg.body = json.dumps({'marathon_minutes_prediction': round(prediction[0], 1)})


Close the editor window. Finally, you just need assign the Docker image to the “Python36 – Inference” operator. As before, right-click the operator and select “Group”. Add the tag “marathontimes”.


Save the changes and go back to the ML Scenario. Now deploy the new pipeline. Select the pipeline “20 Apply REST-API” and click the “Deploy” icon.


Click through the screens until you can select the trained model from a drop-down. Click “Save”.


Wait a few seconds and the pipeline is running!


As long as that pipeline is running, you have the REST-API for inference. So let’s use it! There are plenty of applications you could use to test the REST_API. My personal preference is Postman, hence the following steps are using Postman. You are free to use any other tool of course.

Copy the deployment URL from the above screen. Do not worry, should you receive a message along the lines of “No service at path XXX does not exist”. This URL is not yet complete, hence the error might come up should you try to call the URL.

Now open Postman and enter the Deployment URL as request URL. Extend the URL with v1/uploadjson/. Change the request type from “GET” to “POST”.


Go to the “Authorization”-tab, select “Basic Auth” and enter your user name and password for SAP Data Intelligence. The user name starts with your tenant’s name, followed by a backslash and your actual user name.


Go to the “Headers”-tab and enter the key “X-Requested-With” with value “XMLHttpRequest”.


Finally, pass the input data to the REST-API. Select the “Body”-tab, choose “raw” and enter this JSON syntax:

    "half_marathon_minutes": 120


Press “Send” and you should see the prediction that comes from SAP Data Intelligence! Should you run a half-marathon in 2 hours, the model estimates a marathon time of under 2 hours 24 minutes. Try the REST-API with different values to see how the predictions change. Just don’t extrapolate, stay within the half-marathon times of the training data.



You have used Python in Jupyter Notebooks to train a Machine Learning model and the code has been deployed in a graphical pipeline for productive use. Your business processes and application can now leverage these predictions. The business users who benefit from these predictions, are typically not exposed to the underlying technologies. They should just receive the information they need, where and when they need it.

Let me give an example that is more business related than running times. Here is a screenshot of a Conversational AI chatbot, which uses such a REST-API from SAP Data Intelligence to estimate the price of a vehicle. The end users is having a chat with a bot and just needs to provide the relevant information about a car to receive a price estimate.

Happy predicting!




You must be Logged on to comment or reply to a post.
  • SAP Data Intelligence also comes with a built-in Data Lake. If you would like to use that embedded Data Lake instead of AWS S3 you need to do the following changes:

    Section „Data Connection“: Instead of defining the connection to S3, use the pre-defined connection for the DI Data Lake. In the Metadata Explorer go via “Browse Connections” to “DI_DATA_LAKE” and directly upload the CSV file (e.g. in folder “shared”).

    Section “Data exploration and free-style Data Science”: In the Juypter Notebook omit the code for connecting to S3, but instead use the following code to connect to the DI Data Lake and read the CSV file.

    !pip install hdfs
    from hdfs import InsecureClient
    client = InsecureClient('http://datalake:50070')
    with'/shared/RunningTimes.csv', encoding = 'utf-8') as reader:
        df = pd.read_csv(reader, sep=";")

    The access to the built-in data lake from within Jupyter is also described in the documentation.

    Section “Training Pipeline”: Instead of configuring the “Read File” operator for S3, configure the parameter “Service” and set it to “SDL”. Furthermore, adjust the input file path to the execution step in the “Training Pipeline” section, i.e. inputFilePath = shared/RunningTimes.csv

    That’s all you need to do.


    • For the Semantic Data Lake soluton I also had to add the complete input file path to the Execution step in the "Training Pipeline" section.

      i.e. inputFilePath = shared/RunningTimes.csv


      Thanks for the great tutorial and comments!

    • Hi Philip,

      Do we need to do any setting if we wish to use DI_DATA_LAKE for the first time.Its a newly created instance.

      I am getting following error while accessing folders in Data lake for uploading the file. Can you please advise.

      Folder cannot be displayed Error displaying Connection: DI_DATA_LAKE, Path: /. Select a connection or go back to previous location

      Could not browse the connection or folder
      Could not browse the connection. Invalid input.
      Error at Type Extension ‘onBeforeSendResponse’: Error at exchange with Storage Gateway: 404 – “<html>\n<head>\n<meta http-equiv=\”Content-Type\” content=\”text/html;charset=ISO-8859-1\”/>\n<title>Error 404 </title>\n</head>\n<body>\n<h2>HTTP ERROR: 404</h2>\n<p>Problem accessing /webhdfs/v1/. Reason:\n<pre> Not Found</pre></p>\n<hr />\n</body>\n</html>\n”
    • I would add that if you can't upload the file using the Metadata Explorer (the upload button is grey) it could be that is because you don't have the sufficient permissions. Your user must have the policy sap.dh.metadata assigned to them.

      In order to assign it you can go to System Management > Users then click on the plus sign and add that policy to the user.


      Thanks for the tutorial!


    • It’s worth noting that the Data Intelligence virtual machines (DI VMs) do not have the DI_DATA_LAKE connection as of this writing; therefore I had to use an S3 bucket.

      Within the “10 Train” pipeline I also had to make the following changes to the Artifact Producer:

      • Right-click the Artifact Producer operator, and click “Open Operator Editor”.  Click the “Configuration” tab, and change the “service” parameter’s value from “SDL” to “File”.
      • Open the script for Artifact Producer, and change the value of the “flagUsingFileStorage” variable to “true”.

      On a related note, I found “kubectl get pods -w” and “kubectl log {pod_id}” to be invaluable for diagnosing why my pipeline execution kept failing.  These logs can also be downloaded from the Status tab in the pipeline modeler.

      EDIT:  Actually, if you just click the name of the item in Status, you get something more readable than the log, and you can see exactly where in the pipeline the execution failed. 

      • Hi Prashant,

        I also had the problem in DI_DATA_LAKE, so I followed the steps mentioned above. I

        am getting error in artifact producer stating “runtimeStorage via $VSYSTEM_RUNTIME_STORE_

        MOUNT_PATH not found”.


        PS : While debugging I found the error caused due to the code (if condition) in artifact producer  :

        var ( //deprecated
        	flagCheckRuntimeStorage = true
        	flagUsingFileStorage    = true
        	gRuntimeStoragePath     = "/tenant/runtime"
        if flagCheckRuntimeStorage && flagUsingFileStorage {
        		runtimeStoragePath, ok := os.LookupEnv("VSYSTEM_RUNTIME_STORE_MOUNT_PATH")
        		if !ok {
        			ProcessErrorSetup("runtimeStorage", errors.New("runtimeStorage via $VSYSTEM_RUNTIME_STORE_MOUNT_PATH not found"))
        		if _, err := os.Stat(runtimeStoragePath); os.IsNotExist(err) {
        			ProcessErrorSetup("runtimeStoragePath", err)
        		Logf("ArtifactProducer: using the following runtimeStorage: %q", runtimeStoragePath)


        Please help me to move forward!


        • Hi Raevanth,

          I actually haven't done anything further with Data Intelligence since March, so it's quite likely you're on a newer version.  Maybe my solution no longer works.  Are you using an S3 bucket?

          I'm not actually on the DI team, so take my suggestion with a grain of salt, but I suppose one "hacky" fix would be to just change flagCheckRuntimeStorage to false.  Maybe this will break something else though.

          If you'd like I can inquire further.  I have a colleague who has done more with DI than I have.  However you may want to try a more "official" DI support channel first.  I'm sure when you started your DI trial there was some support information provided.

          By the way, was your DI_DATA_LAKE problem this by any chance?  Just happened to stumble onto it.


    • Hi Philipp,

      Could you please also help me with how to export a CSV file to the data lake? My code is like below:

      result = pd.merge(dataset1, dataset, left_on = 'Income', right_on = [0])
      !pip install hdfs
      from hdfs import InsecureClient
      client = InsecureClient('http://datalake:50070')
      with client.write('/shared/DEMO_15761/CustomerSegmentation_Results.csv', encoding = 'utf-8') as writer:

      I got an error saying that “Invalid file path or buffer object type: <class ‘hdfs.util.AsyncWriter’>”

      Thanks for any hint.


  • Thnx Andreas for this great tutorial.

    I got error 502 “bad gateway” when sending the POST request.

    It could be fixed by extending the Python Consumer template:
    when returning the response a new instance of api.message was created.

    request_id = msg.attributes['']
    response = api.Message(attributes={'': request_id}, body=msg.body)
    api.send('output', response)

  • What all needs to be changed to train a different type of model using the pipeline interface? I trained a regression tree model ( successfully in the Jupyter Notebook interface, then transferred to code to the Python3 operator in the pipeline interface. That operator seems to run fine, but the pipeline gets stuck on the "Artifact Producer" operator and the model is not produced (however, the metrics are produced). The pipeline stays in "running" status and never reaches "completed".

    • Working with SAP support we found that there is a 10MB size limit on objects to fit through the pipeline between the Python3 operator and the Artifact Producer operator. In my case the model was larger and was getting stuck. We solved this by grouping those two operators together, tagged with the dockerfile below. Note: some of these settings are specific to my instance of DI


      FROM §/
      RUN python3.6 -m pip --no-cache-dir install \
      requests==2.18.4 \
      tornado==5.0.2 \
      minio==3.0.3 \
      hdfs==2.5.0 \
      Pillow==6.0.0 \
      tensorflow==1.13.1 \
      numpy==1.16.4 \
      pypng==0.0.19 \
      pandas==0.24.0 \
      • You don't need to have all operators running the same image/dockerfile to be able to run them in the same pod. A pod can have several containers based on different images. In DH, the way you tell it to run all operators in the same pod is by adding them all to the same group.

  • Very interesting and complete blog Andreas Forster, thank you!

    I am trying to do the same having loaded the marathon times file on a HANA table.

    In the Jupyter notebook all good, I have used the new python APIs, but then? How should I change the training pipeline?




    • Hello Elisa, thank you for the feedback! To follow this tutorial with data that is held in SAP HAN you need to replace the "Read File" operator in the training pipeline with a "HANA Client". The select statement that retrieves the history can be passed to the HANA Client with a "Constant Operator". And a "Format Converter" after the HANA Client might be needed to get the CSV format this tutorial is based on.
      However, with data in HANA, there is an alternative, that I would prefer. Since in HANA we have predictive algorithms, you can train the model directly in SAP HANA without data extraction. We have a new Python and R wrappers that make this pretty easy. Here are some links on that wrapper
      There are also hands-on sessions on that wrapper at this year's Teched season that is just starting, ie DAT260, SAP Data Intelligence: Machine Learning Push-Down to SAP HANA with Python

        • The only documentation I am aware of is the popup within Jupyer. Have the cursor within the round brackets of NotebookConnectionContext(connectionId='YOURCONNECTION') and hit SHIFT+TAB.

      • Hi Andreas,


        also thanks for your great blog - I'm also actually trying to access HANA from a canary service instance but cannot connect (ESOCKETTIMEDOUT). Is there something specific to configure?




        • Hi Rudolf, I have seen the ESOCKETTIMEDOUT error when
          - the HANA system is either down
          - or if the HANA system cannot be found (ie if it has no external IP address)
          - or if the wrong port was specified

          In case this doesn't help, please ping me directly so we keep the comments on this blog close to the Marathon case 🙂

          Many Greetings, Andreas

      • Hello Andreas,

        It's not very clear how to use Constant Operator in combination with Format Converter, HANA Client operator and Artifact Producer. Will you be able to share how the pipeline would look if we use HANA Client Operator to read the HANA table instead of the Read File operator?


        Achin Kimtee


        • Hi Achin, To change the data source from CSV to SAP HANA in the training pipeline, the python code just needs to be adjusted slightly as SAP HANA is providing the data in JSON format (not CSV).

          import json
          df_data =  pd.read_json(io.StringIO(data))

          The full code is then:

          import json
          import io
          import pandas as pd
          def on_input(data):
              # Obtain data
              df_data =  pd.read_json(io.StringIO(data))
              # Get predictor and target
              x = df_data[["HALFMARATHON_MINUTES"]]
              y_true = df_data["MARATHON_MINUTES"]
              # Train regression
              from sklearn.linear_model import LinearRegression
              lm = LinearRegression()
    , y_true)
              # Model quality
              import numpy as np
              y_pred = lm.predict(x)
              mse = np.mean((y_pred - y_true)**2)
              rmse = np.sqrt(mse)
              rmse = round(rmse, 2)
              # to send metrics to the Submit Metrics operator, create a Python dictionary of key-value pairs
              metrics_dict = {"RMSE": str(rmse), "n": str(len(df_data))}
              # send the metrics to the output port - Submit Metrics operator will use this to persist the metrics 
              api.send("metrics", api.Message(metrics_dict))
              # create & send the model blob to the output port - Artifact Producer operator will use this to persist the model and create an artifact ID
              import pickle
              model_blob = pickle.dumps(lm)
              api.send("modelBlob", model_blob)
          api.set_port_callback("input", on_input)


          That's the pipeline.


          The Constant Generator is just sending the SELECT to the HANA Client, ie

          Just to mention, that if the data is in SAP HANA, I suggest to try using the ML algorithms in HANA to avoid the data extraction (Predictive Algorithm Library, Automated Predictive Algorithm Library). Here is an example




          • Hi Andreas,

            Thanks for your quick and detailed response. It helped me to move in the right direction. There are 2 other issues which I am facing when I design the pipeline as you suggested.

            1. I used constant generator and put select query in the content but in the output of first SAP HANA Client operator, ordering of columns is randomly changing. It’s not the same sequence as present in the HANA table. As ordering changes randomly, python operator can’t accept the same and consequently, pipeline fails.
            2. Float values are coming as 52456/1000 instead of 52.456.

            Did you come across any such issues in your pipeline? Any inputs are welcome.


            Training Pipeline using HANA Table as Source



            Achin Kimtee

          • Hi Achin, You can control the column sequence by replacing the * with the column names.

            I am surprised though that Python is expecting the columns in a certain order, since the code is retrieving the columns by name, not by index?
            y_true = df_data["MARATHON_MINUTES"]

          • Hi Andreas Forster ,

            Even though I give column names instead of select *, the columns does not come in order.

            But you are right. Even though the columns are not in order,the code accepted it and ran successfully. I could successfully generate model using HANA table instead of file.


            Achin Kimtee

    • Hi Elisa,

      I am trying to design a similar training pipeline where input data is coming from HANA table. I get an error if I try to use HANA Client Operator with Artifact Producer. If I use Read File Operator, it works fine. Did it work for you? Can you share how did you achieve this scenario?



      Achin Kimtee

  • Hello,

    When I call the rest-API via Postman I get

    Is there something wrong or do I just have to wait - actually I've been waiting already a rather long time? The inference pipeline in the modeller is running.

    Thx, Ingo

    • Hi,

      I think the reason might be that the Semantic Data Lake (SDL) does not really work in the CAL environment where I tried to deploy the solution. In particular it is possible to save the model in the SDL which can be checked in the Metadata Explorer. However trying to access the model when using it in the inference operator seems to fail. You can also find the error message in the else-branch of the interference operator.

      Instead I used the local file system of DH to store the model which worked fine.

      Regards, Ingo

  • Hello Andreas,


    Great Blog. I have one further question.


    Is it possible to access Vora from JupyterLab Notebooks to access multiple Parquet files and leverage SQL on files.


    Thanks, Prem


    • Hi Prem, I haven't tried accessing Vora from Python. But you can read Parquet files in Jupyter Notebooks and in the graphical pipelines, the latter of which can also write to Vora if that's any helpful. Greetings, Andreas

  • Hi, Andreas,


    thanks for the great hands-on tutorial.


    But somehow I am stuck with the docker build step. Using your sample code, I get the error message:

    Error reading Dockerfile open /vrep/vflow/dockerfiles/com/sap/opensuse/python36/Dockerfile: no such file or directory


    Trying to switch to other base images such as §/ or  python-3.6.3-slim-strech, I am able to build the docker image. Nevertheless, the execution of the pipeline always failed with no further message than "Dead".


    -- Is there any documentation about  base images available for data hub? Is there any way to take a look at the repository?

    -- How to get more detailed error messages?

    • Hi DaPeng,
      The newly release DI 3.0 introduced some changes, there is a new base image for instance.
      I hope to test this out soon, in the meantime please try this syntax

      FROM $
      RUN python3.6 -m pip install numpy==1.16.4 --user
      RUN python3.6 -m pip install pandas==0.24.0 --user
      RUN python3.6 -m pip install sklearn --user

      The Monitoring tile on the DI launchpad shows a more detailed error for a failed pipeline.

      • Hi Andreas,

        I have the same issue with Dapeng, after trying your new script. it gives me an error :

        "Unable to start docker file build: /service/v1/dockerenv/deploy/python36marathon"


        "build failed for image:"

        I checked also the guide in, couldn't figure out how to correct the script.

        Can you please help?

        Thank you!

        • Hi DaPeng, hi James,
          I have just updated the whole blog for DI 3.0.
          There were plenty of changes, the product is moving fast 🙂
          For instance the connection, the docker file, the read file operator the versioning.
          The docker syntax is updated in the blog. And the tags that need to be assigned have also changed.
          You may want to skim through the blog from the beginning, for instance the bucket name is now specified in the connection.
          Please let me know if you get stuck somewhere or if you make it to the end.

  • Hi,

    I am able to build the docker image with following DOCKERFILE.

    FROM $
    RUN pip3.6 install –user numpy==1.16.4
    RUN pip3.6 install –user pandas==0.24.0
    RUN pip3.6 install –user sklearn


    Nevertheless, when running the pipeline, I got the following error:

    container has runAsNonRoot and image will run as root


    Any idea, how to configure the security or runAs context for the container?


    • Hi DaPeng, hi James,
      I have just updated the whole blog for DI 3.0.
      There were plenty of changes, the product is moving fast 🙂
      For instance the connection, the docker file, the read file operator the versioning.
      The docker syntax is updated in the blog. And the tags that need to be assigned have also changed.
      You may want to skim through the blog from the beginning, for instance the bucket name is now specified in the connection.
      Please let me know if you get stuck somewhere or if you make it to the end.

      • Hi Andreas,

        many thanks for this interesting post,

        I did install SAP DI trial version and I'm using AWS for the as a host,

        I used this post to simulate my ML model of predicting diabetes (classification problem), however, I stuck while creating the environment of python through the docker file.

        can you please have a look into the below error?

        many thanks for your help


        Unable to start docker file build: /service/v1/dockerenv/deploy/dockerized

        • Hello Mohamed,
          I wonder if your environment might not be configured correctly, were you able to build any Docker File?
          Does this single line build successfully for example?

          FROM $


          • Hi Andreas,

            I tried this line as well and it runs successfully,

            FYI, I used another dataset, however, I'm using only Pandas, Sklearn, and NumPy to import



          • Hi Mohamed, If the single FROM line builds successfully it seems to find the base image. Can you also build the following code?

            In case this fails, I assume there is a problem with the Docker configuration, maybe it cannot connect to the web to download the packages.

            FROM $
            RUN pip3.6 install --user numpy==1.16.4
            RUN pip3.6 install --user pandas==0.24.0
            RUN pip3.6 install --user sklearn
          • Hi Andreas,

            I tried the above but it gives me the below error


            error building docker image. Docker daemon error: The command '/bin/sh -c pip3.6 install --user numpy==1.16.4' returned a non-zero code: 1
            a few seconds ago
          • Hello Mohamed, I assume your Docker environment might not be configured correctly.
            Maybe it doesn't have Internet access to download packages from PyPI.
            Can you log a ticket with support for this? Or the wider SAP community might be able to help

          • Hi Andreas/Mohamed,


            I am the same error:


            The command ‘/bin/sh -c pip3.6 install –user numpy==1.16.4’ returned a non-zero code: 1


            With this script

            FROM $
            RUN pip3.6 install --user numpy==1.16.4
            RUN pip3.6 install --user pandas==0.24.0
            RUN pip3.6 install --user sklearn
            RUN pip3.6 install --user tensorflow

            I tried to do it like this but the docker is never created

            FROM §/ python3.6 -m pip --no-cache-dir install \
            tornado==5.0.2 \
            hdfs==2.5.0 \
            tensorflow==1.13.1 \
            numpy==1.16.4 \


            I think that is not posible create docker file into Sap Cal DI 3.0. I tried with Azure and i am triying it with Aws.


            Therofere when i assign tags to groups i get the error:

            failed to prepare graph description: failed to select image: no matching dockerfile found for group 'group1' with runtime tags: {"pandas": "", "sklearn": "", "numpy": "", "Pillow": "", "test2": "", "hdfs": "", "tornado": "", "minio": "", "tensorflow": "", "pypng": "", "requests": "", "python36": ""}; Alternatives are: { {"dh-app-base": "2002.1.10", "node": "vflow-sub-node"}} { {"flowagent-codegen": "2002.1.12", "opensuse": "", "spark": "2.4.0", "hadoop": "2.9.0", "python36": "", "tornado": "5.0.2", "deprecated": ""}} { {"opensuse": "", "python36": "", "tornado": "5.0.2", "sapgolang": "1.13.5-bin", "zypper": "", "deprecated": ""}} { {"ml-python": "", "numpy36": "1.13.1", "scipy": "1.1.0", "tornado": "5.0.2", "sapgolang": "1.13.5-bin", "python36": "", "opencv36": "3.4.2", "pykalman36": "", "textblob36": "0.12.0", "tweepy36": "3.7.0", "automated-analytics": "", "sles": ""}} {test2 {"test2": "", "sles": "", "python36": "", "tornado": "5.0.2", "node": "10.16.0"}} { {"opensuse": "", "deprecated": "", "flowagent": "2002.1.12"}} { {"sles": "", "sapjvm": "", "python36": "", "tornado": "5.0.2"}} { {"vflow_textanalysis": "", "python36": "", "tornado": "5.0.2", "sles": ""}} {org.opensuse {"tornado": "5.0.2", "sapgolang": "1.13.5-bin", "zypper": "", "opensuse": "", "python36": ""}} {phyt {"requests": "", "tornado": "", "tensorflow": "", "sklearn": "", "pypng": "", "pandas": "", "minio": "", "hdfs": "", "Pillow": "", "numpy": ""}} { {"python36": "", "pandas": "0.20.1", "tornado": "5.0.2", "node": "10.16.0", "sles": ""}} { {"sles": "", "vflow_dh_dq": "2002.1.2"}} { {"default": "", "sles": "", "python36": "", "tornado": "5.0.2", "node": "10.16.0"}} { {"dh-app-data": "2002.1.9", "node": "vflow-sub-node"}} { {"sapgolang": "1.13.5-bin", "deprecated": "", "node": "10.16.0", "tornado": "5.0.2", "python36": "", "requests": "2.22.0", "sles": ""}} { {"sles": "", "sapjvm": "", "streaming_lite": "", "python36": "", "tornado": "5.0.2"}} { {"sles": "", "flowagent": "2002.1.12"}} {python36marathon {"marathontimes12": "", "sles": "", "python36": "", "tornado": "5.0.2", "node": "10.16.0"}} {trey {"trt": ""}} {Test7 {"sles": "", "python36": "", "tornado": "5.0.2", "node": "10.16.0", "tes7": ""}} { {"hdfs": "2.5.0", "aiohttp": "3.5.4", "sapdi": "0.3.23", "hana_ml": "1.0.8.post5", "sapgolang": "1.13.5-bin", "hdbcli": "2.4.167", "pandas": "0.24.2", "tornado": "5.0.2", "python36": "", "sles": "", "requests": "2.22.0", "backoff": "1.8.0", "uvloop": "0.12.2"}} { {"node": "", "papaparse": "4.1.2", "sles": "", "python36": "", "tornado": "5.0.2"}} { {"sapautoml": "2.0.0", "tornado": "5.0.2", "python36": "", "sles": ""}} { {"hadoop": "2.9.0", "python36": "", "tornado": "5.0.2", "flowagent-codegen": "2002.1.12", "sles": "", "spark": "2.4.0"}} { {"sles": "", "sapgolang": "1.13.5-bin", "python36": "", "tornado": "5.0.2"}} { {"sles": "", "sapjvm": "", "python36": "", "tornado": "5.0.2", "hanareplication": "0.0.101"}} {test1 {"sles": "", "python36": "", "tornado": "5.0.2", "node": "10.16.0"}} {Tes6 {"Test6": ""}} { {"opensuse": "", "deprecated": "", "vflow_dh_dq": "2002.1.2"}} { {"rjsonlite": "", "tree": "", "sles": "", "python36": "", "opensuse": "", "rserve": "", "tornado": "5.0.2", "r": "3.5.0", "rmsgpack": ""}} { {"tornado": "5.0.2", "sles": "", "node": "", "python36": ""}}
            But i assign the tags. I assign the tags to the operator and the group.

            Great blog Andreas.

          • Hi Andreas,

            I'm having exactly the same problem as Sergio. I was checking other blogs but I think there is no workaround to solve this issue. Could you please help us?

            Truly thanks in advance.

          • Hi Andres Levano , hi Sergio Peña , hi Mohamed ,
            I am bit puzzled that the Docker code doesn't build for you.
            Maybe your systems have something in common, that prevents this.
            I wouldnt be surprised for example, if your Docker cannot connect to the web, and therefore cannot download packages.

            Can you please test whether you can build the following code from the documentation.

            DI Cloud
            FROM $
            RUN pip3.6 install --user numpy=="1.16.1"

            DI on premise
            FROM $
            RUN pip3.6 install --user numpy


          • Hi Andreas Forster ,

            tried again the above, but still facing the below error.

            can you please advise if there is a way to reinitialize the DI to have a fresh copy or do I need to reinstall anything again? as I think there might be an issue with the installation.



            error building docker image. Docker daemon error: The command '/bin/sh -c pip3.6 install –user numpy==”1.16.1″' returned a non-zero code: 1

          • Hi Mohamed , Andres Levano , Sergio Peña , Klaus-Dieter Steffen
            To test this out I have created my on DI trial instance and can now reproduce the error. The Docker code from the documentation, which builds fine in the full DI version,

            FROM $
            RUN pip3.6 install --user numpy

            is giving this error in the trial:
            error building docker image. Docker daemon error: The command ‘/bin/sh -c pip3.6 install –user numpy’ returned a non-zero code: 1

            The following Docker build succeeds in the DI Trial. However, I am not sure if that image is intended to be used. So far I have not managed at least to also install pandas and sklearn

            FROM $
            RUN python3.6 -m pip install numpy

            Let me check with the colleagues that look after the DI trial.

          • Hi Andreas,


            I test it and i get other error. I check my conexion but it is fine.


            [object Object],DEBUG,"[stream] [91mNo matching distribution found for tensorflow
            [object Object],DEBUG,"[stream] Removing intermediate container 35cdef30f81e",vflow,container,35605,func1
            [object Object],DEBUG,"[error] The command '/bin/sh -c python3.6 -m pip install tensorflow' returned a non-zero code: 1",vflow,container,35605,func1
            [object Object],ERROR,"Error building docker image: The command '/bin/sh -c python3.6 -m pip install tensorflow' returned a non-zero code: 1",vflow,container,35269,buildImageCoreDocker
          • Hi Mohamed  , Andres Levano  Sergio Peña , Klaus-Dieter Steffen ,
            Dimitri Vorobiev was able to analyse this behaviour further and identified that the issue is Docker on AWS not being able to connect to the web. Hence Docker cannot download / install the required Python packages. He is also proposing a workaround, please see


  • Hi Andreas !


    This is a great post !! Thanks for sharing all details and steps.


    I created a ML Scenario, in the Jupyter Notebook i installed opencv-python


    !pip install opencv-python


    And everything looks fine


    Collecting opencv-python
      Downloading (28.2MB)
         |████████████████████████████████| 28.2MB 12.6MB/s eta 0:00:01    |████████████                    | 10.5MB 12.6MB/s eta 0:00:02
    Requirement already satisfied: numpy>=1.14.5 in /opt/conda/lib/python3.7/site-packages (from opencv-python) (1.18.1)
    Installing collected packages: opencv-python
    Successfully installed opencv-python-
    When I try to import cv2 I get the error below
    import cv2
    ImportError                               Traceback (most recent call last)
    <ipython-input-3-c8ec22b3e787> in <module>
    ----> 1 import cv2
    /opt/conda/lib/python3.7/site-packages/cv2/ in <module>
          3 import sys
    ----> 5 from .cv2 import *
          6 from .data import *
    ImportError: cannot open shared object file: No such file or directory

    This looks to me that an os library is missign in the container.

    If I try to run zypper I get the below error

    ERROR: zypper was removed due to licensing reasons.
    It depends on 'rpm' module, which is licensed under sleepycat license.
    Do you have any hints to run opencv in ML Scenarios?
    Best regards,


  • Hi Andreas,

    thank you for the comprehensive article. I've met however an issue on the last step. I set a POST request using curl and get an error:

    ralf@home ~> curl -X POST --data "{\"half_marathon_minutes\"
    : 120}" --user 'default\XXXXXX:XXXXXXXX' -H "X-Requested-With: XMLHTTPRequest"

    Illegal URL request path '/openapi/service/f29ab031-5823-479a-b69f-8076c46921fe/v1/uploadjson'. No executing host found

    ralf@home ~>

    The URL is copied from deployment page. Google didn't help. What could it be?



    • Hi Mychajlo,
      Does the inference pipeline still show as "Running" in DI? Maybe it terminated with an error.
      If it is still running, I wonder if your corproate network or security settings might be blocking some communication.
      Can you please try from your laptop using a Hotspot from your phone, or just from your personal laptop with your home Wifi.
      Since I haven't called the inference URL from curl, do you get the same result in Postman? Just in case curl might need some additional config / parameter.

  • I am following this tutorial. I am executing the code below in the jupyter notebook with my own csv file located as in the code:

    import boto3
    import pandas as pd
    import io
    client = boto3.client('s3', aws_access_key_id=access_key, aws_secret_access_key=secret_key)
    bucket = 'bucket1'
    object_key = 'bucket1/shared/JR/RunningTimes.csv'
    csv_obj = client.get_object(Bucket=bucket, Key=object_key)
    body = csv_obj['Body']
    csv_string ='utf-8')
    df_data = pd.read_csv(io.StringIO(csv_string), sep=";")


    I am getting the following error: Can someone let me know what I am doing wrong?

    ClientError                               Traceback (most recent call last)
    <ipython-input-30-e77a0f1f08d7> in <module>
          5 bucket = 'bucket1'
          6 object_key = 'RunningTimes.csv'
    ----> 7 csv_obj = client.get_object(Bucket=bucket, Key=object_key)
          8 body = csv_obj['Body']
          9 csv_string ='utf-8')
    /opt/conda/lib/python3.7/site-packages/botocore/ in _api_call(self, *args, **kwargs)
        314                     "%s() only accepts keyword arguments." % py_operation_name)
        315             # The "self" in this scope is referring to the BaseClient.
    --> 316             return self._make_api_call(operation_name, kwargs)
        318         _api_call.__name__ = str(py_operation_name)
    /opt/conda/lib/python3.7/site-packages/botocore/ in _make_api_call(self, operation_name, api_params)
        633             error_code = parsed_response.get("Error", {}).get("Code")
        634             error_class = self.exceptions.from_code(error_code)
    --> 635             raise error_class(parsed_response, operation_name)
        636         else:
        637             return parsed_response
    ClientError: An error occurred (InvalidAccessKeyId) when calling the GetObject operation: The AWS Access Key Id you provided does not exist in our records.
    Thanks you for your help.
    • Hello Jude, You are getting this error as you are working with a SAP internal system that does not have a true S3 bucket connected.
      Your boto3 library connects to Amazon, which does not know about the S3 simulator in your system. Hence the message from Amazon: "The AWS Access Key Id you provided does not exist in our records."
      To proceed with the Python code, you can create an S3 bucket on Amazon, and use its credentials in DI.

      • Hi Andreas, that's a great information. I had no idea about that. Thank you very much for letting me know of this.


  • Hi Andreas

    First of all, thanks for this blog, very little information is available on DI and this helps a lot

    In my DI, I have setup an s3 connection and it gives me an "ok" value during connection check. In fact, I am able to connect to aws s3, get the data and run the ML scenario in Jupyter notebook in ML Scenario manager. However, when I try to train the model, the read file gives me a blank screen while trying to get a file, manually keying in the file name does not help and as a result the training job abends.



    Any pointers would be really helpful



    • Hello Vinieth, You can get a more detailed error message in the Monitoring tile on the DI launchpad.
      There might be a tidy up process. So in case your dead pipleline does not show anymore, just produce the error once more and it should show up.

  • Hi Andreas,

    very interesting and well-written post!

    I was able to create the Python Rest API and tested with Postman successfully.

    However, if I try to access the same API from inside a SAPUI5 application (with authorization included), I get access error/failure (403). My research points to CORS restricting access to the remote server (Data Intelligence).

    Did you run into issues like this or have any thought on how to overcome it?

    SAPUI5 app and DI tenant are not on the same server/origin.



    • Hi Alex, I love that you want to extend this and put a UI5 front-end on top! That would make for a great hands-on tutorial. I haven't worked much with UI5 though, but because of CORS Postman requires the X-Requested-With header with value XMLHttpRequest.
      Maybe this can also be added to the UI5 app?

      • Hi Andreas,

        I have tried this, still getting 403 (Forbidden) error in console when accessing the API via a button.

        onPress: function(oEvent) {
        			var oView = this.getView();
        			var oResourceBundle = oView.getModel("i18n").getResourceBundle();
        			var url = oResourceBundle.getText("url").toString().trim();
        				type: "POST",
        				url: url,
        				dataType: "json",
        				// beforeSend: function(xhr) {
        				// 	xhr.setRequestHeader("Authorization", "Basic Z******=");
        				// },
        				headers: {
        					"X-Requested-With": "XMLHttpRequest",
        					"Authorization": "Basic Z******="
        				success: function(response) {
        				error: function(response) {

        This is the ajax call I am making from my SAPUI5 app. As far as I could find, it's how it should be done, but I am open to suggestions of course 🙂

    • Hi Alexander, Hi Andreas,

      have you further investigated this topic?

      I was trying to create a very simple web UI with some Javascript to consume the REST API of the OpenAPI block directly, but I stumped on the same error.

      The only solution that worked for me was to completely bypass the CORS problem with a proxy in between... but this is not a "clean" solution.

      I would really appreciate if you could share any thought or even breakthrough 🙂

      Thanks and Kind Regards,


      • Hi Emanuele,

        I have not investigated much further, but have received a response from SAP DI Team that it might have been a bug that prevented the API call up to SAPUI5 version 1.6 or so...

        Could not test this because our gateway currently has a SAPUI5 version below this...

        Can you share your solution with the proxy? It may not be clean, but maybe worth a try for prototypes/showcases 🙂

        Thanks and best,


  • Hi Andreas.


    I've been following your tutorial and got to a point where i get an error and can not continue.


    I was creating the dockerfile with the content and the tag that you specified. However, When I click "save" the save and build icons are disabled and when I hover above them I get a "prohibited" red icon. In the logs I can see a "warning" log but the message only says "1". Are you familiar with this error?


    I appreciate your help


  • Can the SAP DI ML Model APIs be authentication using Token as well?

    If Yes, where can we find these token on the SAP DI Application.

    User/Password is not the best way for API AUthentication.

    I use the below to call the API but throws error that invalid authentication.

    In that case I want to understand if the user needs to have any bear min privilege ?

    response = requests.request("POST", url, data=payload, headers=headers,auth=HTTPBasicAuth('user', 'password'))

    Thank you!


    • Hello Rahul,
      The Python Consumer template is designed for basic auth with user name and password
      You could ask on the wider forum whether there are alternatives?

      If you get an invalid authentication, please ensure your user is prefixed by the tenant, ie



  • Hi Andreas,

    Thanks for blog. As I am following the steps in the blog but still I am not able to build the docker file.


    Posted this query quite a while ago & proposed solution still have similar error. I am working on Sap DI 3.0


    Looking forward for the solution.

    • Hello Rachit,
      Frank Schuler posted some answers to your question on the community.
      Could you please try those steps and post the outcome on that forum?
      I will subscribe to the question and will try to help if I have any further idea.

      • Hi Andreas,


        Thanks for the response. I have followed the Frank Schuler suggestion but still I am facing same issue. Here is the error I am facing

        failed to prepare graph description: failed to prepare image: error building docker image. Docker daemon error: The command ‘/bin/sh -c python3.6 -m pip –no-cache-dir install –user pandas’ returned a non-zero code: 1


        I am using SAP DI trail 3.0.

        Will be really helpful, if you can help me to resolve this query.



        Rachit Agarwal

        • Hello Rachit, Let's try to narrow it down. It was working successfully with this code and just one tag of your choosing

          FROM $
          RUN pip3.6 install --user numpy==1.16.4
          RUN pip3.6 install --user pandas==0.24.0
          RUN pip3.6 install --user sklearn

          If the build gives the same error, please try building this single line
          FROM $

          If this succeeeds we know that Docker is working in principle at least
          Then build
          FROM $
          RUN pip3.6 install --user numpy==1.16.4

          Should this fail there is a problem adding additional libraries.
          This could indicate that DI cannot download libraries from the web. Please see this post from  Dimitri Vorobiev

          • Thanks Andreas, for giving a helping end. I think Docker is working fine – as I am able to build by running

            FROM $

            Adding libraries is causing similar error.

            Even I tried to follow Dimitri suggested steps but it’s not helping me out at the moment. After deleting the pipeline I am atleast able to run build the docker but as soon as I tag it while executing the pipeline it fails with same error.

            Can you please let me know the way around to it.



            Rachit Agarwal

          • Hello Rachit,
            The test shows that there is some problem with your Docker environment downloading packages.
            I don't know any steps to troubleshoot this beyond Dimitri's suggestion.
            Since that trial issue is not specific to this blog, I suggest to continue the discussion with the whole community through the separate question you had posted.

  • Hi Andreas Forster ,

    As suggested in this post, I did the API call from Postman and it worked for 2-3 times. But after that it started throwing below error.

    Illegal URL request path '/openapi/service/1381f941-8c02-477b-8679-57a07dcd9752/v1/uploadjson/'. No executing host found

    I have stopped the pipeline, re-deployed it but still throwing the same error. Can you pls help with this error?


    Postman API POST Error



    Achin Kimtee

      • Hi Andreas Forster ,

        After re-deployment of pipeline, I am using the updated URL but the issue is still there. What's weird is, it works sometimes and then it stops suddenly with the error I shared in the previous comment.



        Achin Kimtee


        • Hello Achin Kimtee, If the pipeline is still shown as "Running", I am not sure what could be causing this message. Maybe network connectivity, but I am guessing. Can you please ask to the wider community on
          In case the pipeline stopped with an error, the "Monitoring" tile could show further details, on why it stopped.

  • Hi Andreas Forster

    thanks for the comprehensive and interesting tutorial.

    I am running the Trial Version 3.0 on GCP Hyperscaler and I have an issue in the training pipeline: i'm able to see the final model score, but the Artifact Producer runs in the following error.

    Graph failure: ArtifactProducer: InArtifact: ProduceArtifact: response status from DataLakeEndpoint "502 Bad Gateway": "{\n \"message\": \"Error while executing type extension: Error during exchange with Storage Gateway: 400 - undefined\",\n \"code\": \"10102\",\n \"module\": \"dh-app-connection\",\n \"guid\": \"e9c0e17d-c96a-4da2-8dfe-788a56a6bc89\",\n \"statusCode\": 500,\n \"causes\": [],\n \"hostname\": \"datahub-app-core-f1fc5a8fb5566f4e7e2284-6577c59bf7-txhh5\"\n}"

    Could you please indicate me what might be the problem?

    Thanks in Advance,


    • Hello  Andrea Scalvini , Last month Dimitri Vorobiev made a comment that there "is a bug in our automated deployment of DI Trial Edition when deploying on Google Cloud".

      He is giving some suggestions in his post, which might be relevant for you as well.

      In case that's not resolving it, can you please post this behaviour as question to the wider community

      • It worked, thank you very much (and to Dimitri Vorobiev for the support.

        I add a further information that could be useful for someone else, in the Pipeline 20 Apply REST-API, inside the Python36 - Inference there is this line of code:

        model_ready = False

        To have response from the API, it needs to be changed to:

        model_ready = True

        otherwise it would respond with "Model has not yet reached the input port - try again"

  • Hello everyone!


    First of all great tutorial! I am having problems with the execution of the graph. I am using s3 bucket but i've got this message when execute the graph.

    I would apreceate some help here to complete this example.

    Thanks in Advance,


    • Hello Petar, The s3_files connection is just needed to load the CSV file for training the model.
      The Artifact Producer should only use the built-in DI_DATA_LAKE connection.
      Maybe you changed some configuration of the Artifact Producer, or (if you are using DI on-prem) the DI_DATA_LAKE might not be configured incorrectly?

      • Thank you for the fast forward, if i dont have the DI_DATA_LAKE dont work? Its there anyway to work without the DI_DATA_LAKE, because in our onprem we dont have installed the DI_DATA_LAKE.

        • Yes, you need the DI_DATA_LAKE connection to make this work.
          This can still be setup fully on-prem though. Just passing on a comment from a colleauge that helped another on-prem customer:
          You can use an S3 interface such as rook-Ceph. It is essentially an S3 API wrapper around a block storage that is hosted on-premise. You can either deploy rook directly in Kubernetes, which would then leverage the storage provisioner for Kubernetes, or set up a dedicated Ceph filesystem server. From the SAP Data Intelligence perspective it will be the same as connecting to an Amazon S3 storage bucket, except the hostname will be different.

          I believe that such a setup with rook-Ceph is supported. But please contact support in case you need an official statement for this.