A simple example that illustrates how data links work. Two table operators are connected by a data link. The link connects the output of the top table with the input of the bottom table.
The user decides to remove one of the columns from the first table operator (changing a parameter to the table operator).
The output dataframe generated by the top table operator now only consists of the remaining two attributes (Age and CreditScore). The bottom table operator reacts to this change, and will change visual output (the data shown in the operator) as well as it’s output dataframe.
Note that the system color codes dataframes. This color coding is a unique identifier that encodes a particular dataframe and allows users to track if dataframes are distinct from each other. In the example, the output dataframe of the top operator changes it’s color code once the user removes a column since it is now a different dataframe that is outputted (first a dataframe with 3 columns, after change one with 2 columns).
By breaking data links a user can generate different “versions” of a dataframe output. In the example (a) we connect two table operators. We then break the data link (b), but keep the second table operator. By manipulating the columns in the top table operator we now generate a new dataframe output, however we can still use the “old” one (c).