Preview Data - myantandco/RA-BitnobiPilotJuly2020 GitHub Wiki

The purpose of preview data is to show the user some of the data set to have confidence that each workflow step is doing the correct operation. The other goal is to provide the results of the workflow step quickly as possible to allow for real-time editing of the workflow.

Running a workflow operation on a large dataset can take a long time, but running a small subset can be done quickly, and in just a few seconds in many cases. This is why the Preview contains a small subset of the data.

The Workflow Data Source elements (Importer, SQL, Data Source, Workflow) create the initial Preview data set and the subsequent Operator elements in the workflow operate on this set.

Since the Preview operates on a small dataset, it is possible that when creating a Select condition that the resulting preview set does not contain any rows that match the condition.

To reduce the chances of this, Bitnobi does a random sampling of the data source to populate the Preview data. Currently this is done only by the Workflow elements Importer, Data Source, and Workflow.

If you do encounter this condition, you can go to your data source element and press the "Randomize Sample" button in the Preview titlebar to select a different random set, then refresh your Select condition.

Bitnobi has some features to assist in working with very wide datasets (e.g. 800 columns). In this case the Preview pane will segregate the display into pages of 50 columns. A gear icon will then appear in the upper right hand side of the Preview pane titlebar.

Pressing the gear icon will open up a drop-down list of pages and you can then select which page you wish to display.

The Preview pane also has a filter field. You can type in a string pattern and the Preview pane will then display only those columns matching the pattern.