When comparing data in a visualization, it's necessary to interchange datasets quickly, and easily to get a series of choices to act on.
A new feature in the latest version of Oracle Analytics Cloud, called replace dataset, allows users to intelligently apply (or map) data visualization analyses created using one dataset on to another dataset with only a few clicks and intuitive column mapping.
The data visualization feature in Oracle Analytics allows users to create in-depth and insightful analyses, add sophisticated calculations, build visually compelling infographics, and more. Sometimes, after spending significant time and energy on a project, a user realizes they want to perform a similar analysis on a different dataset, or substitute certain columns within those analyses.
The replace dataset feature lets users capitalize on the effort they have already expended on a project by offering an intelligent tool to swap the dataset or columns with their corresponding points in the new query. This feature can also come in handy in situations where you are asked to analyze one batch of data, and later you want to perform the same analyses on a different batch of the same dataset (perhaps for a different date range, or region, or product).
For demonstration purposes in this blog, we will use a sample project that analyzes quarterly value growth, quarterly indexed growth, and previous-quarter growth for a sample sales dataset. The snapshot below gives you an idea of what this project might look like. (click image to see the change)
In this example, this analysis looks good. However, my manager now asks me to perform the same analysis, only this time for different data that I have, such as expenses. To achieve this goal, I choose the replace dataset feature. Let's walk through the process.
The replace dataset feature is quite easy and intuitive to use. To replace a dataset in an existing project with another dataset, I simply right-click on the dataset and select the "Replace Data Set" option. (Note that the Replace Data Set option is available only for datasets that are not already joined.)
This will open a prompt window, which lets me select or map the columns of an existing dataset to the columns from new dataset. Oracle Analytics identifies the columns that are used in the project (visualizations and custom calculations) and prompts me to map only the used columns to relevant columns in the new dataset. The following photo shows what the prompt window looks like.
Once I click the Replace button, all the existing visualizations will be updated with metric and dimension columns from the new dataset, without any manual intervention. In this case, we are replacing the Sample Order Lines dataset with the Expenses dataset. The screenshot below is how the canvases look with the Expenses dataset (with a filter that removes data for 2016, as we do not have complete data for that year). (click image to see the change)
Another important function of this feature: if you want to replace a column in this dataset with another column in the same dataset, you can use the Replace Data Set option, then select the same dataset, and then map the column to be replaced with a new column in the dataset. In this example, if you would like to replace Business Unit with Expense Type, you can map the Business Unit column to Expense Type column.
To summarize, the replace dataset feature lets users choose a project as a template, then use the same analysis for other datasets with just a few clicks and without having to manually re-create the entire project, saving time and effort.
Watch our short video tutorial, which demonstrates the replace dataset feature.