Name: SAP Data Intelligence ML Scenario Manager (03) - Intro to Scenarios & Notebooks
Uploaded: 2023-9-4
Duration: 440 s

Share feedback

<p>Planned maintenance of Practice Systems</p>

Planned maintenance of Practice Systems

Why create an SAP account?

Enhance your learning experience

SAP Learning

Subscribe now

Learn more

Live Sessions

Hands-on Practice

Certification

My Learning

Ready to start learning?

<div><div>In this third of a nine part video series regarding the SAP Data Intelligence Machine Learning Scenario Manager, Tahir Hussain Babar (Bob) gives an overview of the ML Scenario Manager, and explains each of the key tabs ; datasets, notebooks, pipelines, executions, models and deployments. We'll also look at the basics of using Jupyter Notebooks. <br><br>0:00 - Introduction<br>1:00 - ML Scenario Manager<br>2:25 - Create scenario<br>3:50 - Create Notebook<br>4:15 - Working with Jupyter Notebook<br><br><a href="https://sap.to/6056KrmPe" rel="nofollow noopener noreferrer" target="_blank">https://sap.to/6056KrmPe</a><br><br>Please note:<br>                     This video was originally published on the SAP HANA Academy YouTube Channel                      and therefore may contain references to SAP HANA Academy and YouTube.                      The tech world is always evolving and the SAP HANA Academy YouTube channel                      is set to close on December 31, 2023.                      At that time, references will no longer be valid                      and will gradually be removed when the videos are updated.<br>Also, this video includes automatically generated captions                      and therefore might misrepresent the spoken content or display incorrect product names</div></div>

SAP Data Intelligence ML Scenario Manager  (03) - Intro to Scenarios & Notebooks

Welcome to the SAP HANA Academy. My name's Bob, and in this video series, we'll be looking at SAP Data Intelligence. In this video series, we'll be focusing on using the Machine Learning Scenario Manager. And what we'll cover is how to create predictions within the Machine Learning Scenario Manager and then expose them so that they can be accessed via an API. In this video we're going to be going through an overview of the Machine Learning Scenario Manager and we're going to look at creating scenarios and creating notebooks. In the last video we looked at our data set, which is a CSV file in our semantic data lake. In order to examine this file we used something called the SAP Data Intelligence Metadata Explorer and we also had a brief look at the Connection Manager. Again, we're using connections to our semantic data lake. Now in this video, we're going to be looking at some of the concepts surrounding machine learning. And to do this, we use our Machine Learning Scenario Manager. So within SAP Data Intelligence, we can harness the power of machine learning and gain valuable insights into the data. As a data scientist, the machine learning component supports you from the modeling of the data, with this machine learning scenario manager, to also visualizing and analyzing your data using what's called the Metrics Explorer. Now, we're going to be accessing machine learning functionality programmatically and we can do this via an SDK. The SDK is available to us from a Jupyter Notebook or from a pipeline operator. So in this video, we'll be looking at using a Jupyter Notebook. In a future video, we will be looking at how we can use pipelines within our modeler. The Machine Learning Scenario Manager helps you organize your data science artifacts and to manage your tasks in a central location. Now as a multi-faceted data science application, it's built around the key concept of machine learning scenarios which may contain multiple facets such as data sets, pipelines and also these Jupyter notebooks. What you can see here is our machine learning scenarios. At the moment we've not got any and what we'll do is we'll create one. Now these scenarios will capture all the artifacts that you need to conduct your data science experiments. I'll give the scenario a name and also we can have a description which essentially answers a business question. You'll see that there are six different tabs. Now in regards to the first tab, this is your list of datasets which you can bookmark that are going to be used in this scenario. We're not going to register a dataset. You'll understand why later on when we call the dataset from the Jupyter Notebook and also from the pipeline. There are two types of artifacts, both design time and runtime artifacts. Examples of design time artifacts would be pipelines and Jupyter Notebooks. These Python notebooks are to explore and experiment with your data. And these graphical pipelines are then used to deploy the data scientist's logic into production. Runtime artifacts are things like training runs, models, and model deployments. These executions or training runs show which pipelines were executed. The models are those models that were created by those executed pipelines. And deployments are simply the URL of pipelines that are running and available for inference. And we're going to look at all of these options in the next few videos. If you've worked with Python notebooks before, you can skip to the next video. However, over the next two or three minutes, I'm going to cover the basics. So let's create ourselves a new notebook. And what I'll do is I'll call this 00 notebook intro. And when you click on create, you'll be prompted for the kernel that we're going to use. So we'll select Python 3, and then we'll get our notebook. Now these notebooks are popular with data scientists, so they can contain all the code, including the code's output as well as further documentation. This combination makes it very easy to keep track of the code, and to repeat or modify previous steps. So what we'll do is we'll start off with the first line of code, which is a hello world statement. And we're just going to print this to our screen, a very simple statement, and then we'll just need to execute this statement. The number 1 on the left will indicate that this cell was the first to be executed, and then of course subsequent statements will have different numbering, which is useful if you've got a lot of code. So in order to carry out calculations, we can create simple variables. So here I'm creating a variable, my variable, which is equal to the value of 100. Now, as well as executing via the play button, I can also press shift enter to create this variable. What you'll notice is that nothing has been returned. All we've done is assign the value of 100 to my variable. However, what we can do now is modify this calculation. So, for example, I can say my variable actually equals my variable times 2. Again, we haven't returned anything. If we want to return something, we can of course use the print statement with the output of my variable. So, we can simply do a print, the current value of my variable is as a string, and then we can add the my variable and I'll put it in format string. And again, when I execute, we should see that the value of my variable is 200. Of course, we'll be focusing on calculations, but you can also format your notebooks to add things like comments and headers. So, for example, to add a comment, I can simply write calculating my variable like so. However, when I execute it, because it's not Python code, the cell can't be executed. So to execute a comment, all we need to do is change the option to Markdown, and then when I execute you can see here we've got some form of a header for example. Additionally, these cells can be rearranged. So if you want to add headers or comments and place them in certain parts of the code, you simply can drag them up as I've done here. So that's the basics of Python notepads. So let's actually move onwards to some data science. You could put all of this work in one single Jupyter Notebook, but the advantage of being able to have multiple Jupyter Notebooks means you can better organize your code. So if I go back to my Machine Learning Scenario Manager, here we can create those multiple notebooks. Now, what I'll do is I'll go back to that notebook and what we'll do is we'll save the notebook before we continue to the next part. In the next video, we'll be building models using these Jupyter notebooks and we'll be looking at linear regression.