### Input Data Set Recap for Building Time Series Predictive Scenarios

The training data set contains the past observations (the history) that are used to generate the predictive model and the data and time when the observations were recorded.

- In this data set, the values of the signal variable are known.
- The data set might also contain some influencer variables.
- The past and future values of the influencer variables must be known (at least for the expected forecast horizon).

By analyzing the training data set, Smart Predict generates the time series model.

### Limitations

The training or application input data set must not contain more than 1,000 columns.

While applying the predictive model to an application data set, Smart Predict generates additional columns and the application process can get blocked if the application data set already risks crossing the limit of 1,000 columns.