How to define three inputs and one output using recurrent neural network

I have 740 instance (3 inputs and one output). I need to train this dataset using Recurrent Neural Network but I'm new to it. I followed the chickenpox guide on https://www.mathworks.com/help/deeplearning/examples/time-series-forecasting-using-deep-learning.html but I noticed they have no outputs and not more than one input, it was splitted.

For my own dataset, I have 3 inputs and 1 output. Can anyone please help edit the code attached to fit my architecture and make it run?