To fulfill the specified requirement, we can utilize the “split” node to divide the data into the desired ratio. This node produces two outputs: the first edge consists of 75% of the data, which can be utilized as the training data, while the second edge contains 25% of the data, serving as the test data.