Getting Started with KNIME Analytics: Running the Platform

Data Analytics is a process in which tools convert the raw data into meaningful format. For Structured Data, tools are required to handle the missing values, gain insights, convert to appropriate data formats, and generate charts so we can later use the data for further purposes.

Many Data Analytics tools are available out of which KNIME is most popular.

Introduction to KNIME Analytics Platform

KNIME is a one stop Data Analytics platform in which we can import, clean, manipulate data with single clicks. It is no code platform as it makes use of nodes. Each node is a single entity that has some special functionality. The nodes are connected to one another. In KNIME, we simply import our data, choose nodes, get the output and later on export the workflow. Workflow is the region where we drag and drop the nodes. To get started with KNIME, the first step is installing the software on your machine.

You can refer to Download and Installation of KNIME Analytics Platform for understanding installation.

Navigating the User Interface of KNIME

After installing KNIME from the official website, a launcher dialog box will open in which we need to specify the directory in which the workspace is to be saved. There are two versions of KNIME UI: modern and classic. Now there are some parts that make up KNIME which are as follows:

• Explorer: In this section, we are able to see the workspaces that we have created and we can upload all the necessary files in that section.
• Node Repository: A node is the smallest unit that has some functionality. KNIME provides wide range of modes and the collection of the nodes are present in Node repository.
• Workspace: The empty space where we can drag and drop the nodes and create a workflow.
• Description: Here we can set the aim for our KNIME project. It has also other metadata features like Creation Date, Tags, links and author name.

Creating a New KNIME Workflow

To create a new workflow follow the steps:

1. Under the File option click on New option.
2. There will be two options: Create KNIME Workflow and KNIME Workflow group. Click on the first option.
3. A dialog box will pop up in which we have to specify the name and the workspace directory.
4. Provide with the details and click OK. The Workflow Editor will open up.

1. Importing Data into KNIME

KNIME can deal with many data formats including CSV and Excel. Here we have used Netflix OTT Revenue and Subscribers (CSV File). Now we will use CSV Reader Node and double click on that node. Specify the file location and click on execute. In the bottom section we are able to see the table and also the statistics associated with it. The statistics includes column wise mean, quantile, missing values, unique values etc.

2. Basic Data Manipulation - Data Preprocessing

We can perform basic data manipulation like filter, sort, perform mathematical operations , appending the outputs etc. After reading the data, we need to connect the Reader node with the Data Manipulation nodes.

• Column Filter

This node is present in the Data Manipulation part. It is used to filter the column names, specify the patterns, wildcards etc. There are two sections: Include and Exclude. The columns can be filtered using the arrows. Here we have selected few columns to display our data.

From the above output we can see that after executing the node, we can see the four columns as we have only selected four columns.

• Sort Data

To sort our data in ascending or descending order, we can use Sorter node. Establish a connection between CSV Reader and Sorter. Double click on Sorter node. A dialog box will open. Select on which column the sorting is to be done and click on further advanced sorting options if required. Click on apply and execute.

For instance using the Netflix data we have sorted Global Revenue in ascending order.

• Math Formula

We can also perform mathematical operations on KNIME. We can create new columns by simply providing the mathematical operations. The best part of KNIME is that we do not have to remember the syntax of the Operation as it provides the formula n the Description part itself.

Double click on the Math Formula. A dialog box will open. In the left side column names are present and in the category all mathematical operations are present. Select on any one operation and the syntax will appear in the description part. Provide with the new column name and click on OK. Execute the node. Here we have calculate the sum of Global Revenue.

• GroupBy

In KNIME we can group our columns and perform the aggregate operations like Mean, Median etc. Drag and drop the GroupBy node and establish connection with the imported data. Double click and select the grouping column. For Aggregate click on Manual Aggregation option, choose and click on OK. Execute the node to get the results.

From the below output we can see that we have calculated global revenue based on Date.

• Handling Missing Values

Handling missing values is an important step in Data Manipulation. We can also replace the missing values with any value or we can drop the rows as well. To handle the Missing Values, drag and drop the missing value node. Click on Configuration settings and choose what to do in case a missing vale exists. Then execute the node.

• Joining

Joiner node is used to join two or more tables based on a common column. It is similar to the SQL joins. We can specify what output to display whether the matched rows or left/right unmatched rows. The Joiner node needs two input tables.

As we can see here we have used the date column to join the tables. It returns the matching rows present in both the tables.

Saving and Exporting Workflows in KNIME Analytics

KNIME provides the facility to save and export the workflows. For saving the workflows follow the steps:

1. For modern UI click on the Save button present in the left side.

2. By default the local workspace is defined. Provide a name and click on OK.

The shortcut key to save the workflow is CTRL+S

Exporting Workflows

We can export our workflows as well on KNIME. Exporting means saving our workflows on our local system. The steps are as follows:

1. Switch to the classical UI.
2. Click on File.
3. From the list of options click on Export KNIME Workflow.
4. A dialog box will open. Specify the path where the workflow is to be saved and click on Finish. The workflow will be saved with .knwf extension.

Conclusion

KNIME Analytics Platform is a versatile and powerful tool that simplifies data analytics through its drag-and-drop interface and wide range of built-in functionalities. From data preprocessing to machine learning, KNIME is a great platform for beginners and experts alike. By following the steps outlined in this article, you can start building your own workflows and exploring the vast possibilities of data analysis with KNIME.

baidehi1874

Follow

Improve

Article Tags :

• AI-ML-DS Blogs
• Data Analysis
• AI-ML-DS
• AI-ML-DS With Python