Time series and composite visualizations
1. Time series and composite visualizations
Let’s see how to visualize time series data in KNIME Analytics Platform. This table contains the expenses of the office in the last weeks. You can filter one category, for example, Technology, and show its values in a line plot. The data, however, contain multiple categories that might be interesting to plot, especially as multiple lines to be compared. To achieve that, you need to manipulate the data structure. The goal is to obtain a table with one column for each line that you want to represent. In this case, you can achieve that with a Pivot node, to obtain one column for each category. The plot obtained makes it easy to compare the expenses in the various categories at different times of the year and reveal seasonality. However, other stakeholders, such as the finance department, might be interested in seeing how the expenses sum up towards the end of the year, especially if there is a budget limit. For that, you need a cumulative plot, where the value of each measurement, in this case, the expenses of each week, is summed with the previous ones. There are many ways to achieve that, and one of them is by using the Moving Aggregation node. Select Cumulative computation, then add the column that you want to aggregate and the operation, in this case sum. Now you can see that the plot shows an increasing line, with the values summing up over the course of the year. Make sure to indicate that this is a cumulative plot in the plot title, to not mislead the reader. If you want to add a reference line, for example to represent the maximum budget or the desired value, you can create a new column with a Constant Value Column node and add it to the plotted lines. Now that you have seen the different plots and have created a few of them, it feels natural to group them together in a composite visualization, to see them one next to the other in the same page. In KNIME, composite visualizations are created by components. Select the visualizations that you want to group and right-click create component. The nodes are replaced by a gray node, called component. You can still see what’s inside by opening it. When a component contains visualization nodes, like in this case, all visualizations are collected in the component composite view. You will learn more about composite views in the next chapters. For now, let’s practice what you’ve learned so far!2. Let's practice!
Create Your Free Account
or
By continuing, you accept our Terms of Use, our Privacy Policy and that your data is stored in the USA.