Skip to main content
The Apply Transform step applies prebuilt transformations to the data, and lets you combine several at once.
The options available depend on the frequency of the input data. A daily dataset offers rolling-window transforms; a monthly one offers seasonal adjustment and sum-last-12-months; a yearly one offers neither. Every frequency includes Original Variable, which keeps the untransformed column alongside the transformed ones.
Some transformations need reference data that Alphacast supplies automatically — inflation indices for constant prices, population for per capita, GDP series for % of GDP, and exchange rates for currency conversion. These require the dataset to carry a country entity so the right reference series can be matched. Each transformation adds a new column named after the original plus the transformation’s internal key — EMBI Global Diversified Subindices - 30d_change. The original column is kept, so you can apply several transformations to the same variable and compare them side by side.
For seasonal adjustment with control over the underlying model — including trend and irregular components — use the X13 ARIMA Seasonal Adjustment step rather than the prebuilt Seasonally adjusted transformation.
Country-specific unit conversions are also available on their own through the Apply Country Transform step. If none of the prebuilt options fit, Calculate Variable lets you write the transformation yourself — see the formula reference.

Next steps

Calculate variable

Write custom formulas when a prebuilt transform doesn’t fit.

Change frequency

Resample first to unlock frequency-specific transforms.