TS Model Factory Tool

The TS Model Factory tool estimates time series forecasting models for multiple groups at once using the autoregressive moving average (ARIMA) method or the exponential smoothing (ETS) method. (To generate time series models for a single group, use the ARIMA Tool or ETS Tool, which have more functionality for single groups.) If using the ARIMA method, the tool can also use related covariate fields to make a more accurate prediction.

This tool uses the R tool. Go to Options > Download Predictive Tools and sign in to the Alteryx Downloads and Licenses portal to install R and the packages used by the R Tool. See Download and Use Predictive Tools.

Connect an input

The TS Model Factory tool requires a Designer data stream with at least two fields: the group name, which must be a String, VString, W_String, or V_WString, and the target field, which must be numeric. Covariate fields may also be present, as well as fields that will not be used in model creation.

Configure the tool

Model Configuration

• Type of Time Series Model: Select the method to use to generate the time series model for each group. You must use the same method for all groups; however, you can filter your data into two different groups and use ARIMA on one group and ETS on the other.
• Use covariates in model estimation? (ARIMA only): If you're creating an ARIMA model and wish to use covariates, select this option, and then select the fields you want to use as covariates.
• Select the target field: Select the field you wish to forecast. This field must be numeric and have at least two unique values.
• Select the grouping field: Select the field with the names of the groups.
• Time period type: Select the option that matches the measurement frequency in your data. For example, select Monthly if your data was measured on the 1st of every month. This field also determines the minimum amount of data needed for each group. You need to provide at least two full repetitions for each group. For example, if you select Hourly, you would need to have at least 48 measurements per group (24 hours in a day * 2 = 48).
• Series starting period (optional): Select this option to specify when the target series starts. For example, if your time period is monthly and your series starts on April 1, 2013, you would choose "2013" for the years the series starts and "4" for the week, month (numeric), or quarter of the series start.

View the output

Connect a Browse tool to each output anchor to view results.

• O anchor: Consists of an output stream containing the ARIMA or ETS model object that can be used for both point forecasts and a user-specified percentile confidence interval surrounding those forecasts.
• R anchor: Consists of a table with information about various statistical measures and model parameters.