This workflow performs a time series prediction using a recursive loop. In the workflow the first metanode generates some time series data for the saw tooth wave to train a linear regression model. It then uses the model in a recursive loop recursive loop to predict the values for the next 30 time steps. One time step per iteration.