Charts on a dataset and in a visual analysis¶
The DSS data visualization component lets you create a large variety of charts with a simple drag-and-drop interface.
The charts component is available in two places in DSS:
- At the dataset level: when you are in a dataset, click the Charts tab.
- At the visual analysis level, you will also find a Charts tab.
While the basic charting features are the same in both usages, there are several key differences.
- Charts on datasets are based on the dataset itself.
- Charts in a visual analysis are based on the result of the preparation script of the analysis.
In both cases, charts are by default based on a sample, but this sample can be configured to extend to the whole dataset. The sample can always be configured in the Sampling tab in the left column.
In a visual analysis, the sample is computed before applying the preparation. If you have a sample made of 20,000 rows but your preparation removes 4,000 rows, the chart will be based on 16,000 rows.
Charts in a visual analysis cannot use the “Live” processing engine which runs computations in the database.
For more information, see Sampling and charts engines.
Charts in a visual analysis cannot be published on a Dashboard. Only charts on a dataset can be published.
If you want to publish charts in a visual analysis, you need to first Deploy your analysis to make a Prepare recipe and a new dataset. When you Deploy, you have an option to copy the charts that were on the visual analysis in the output dataset.