apply_official_tcode – Apply McCracken-Ng’s series-by-series stationarity transforms.#

Back to transform_policy axis | Back to L2 | Browse all options

Operational op under axis transform_policy, sub-layer l2_b, layer l2. Standalone callable: mf.functions.apply_tcode_transform.

Function signature#

mf.functions.apply_tcode_transform(
    panel: pd.DataFrame,
    tcode_map: dict[str, int],
) -> pd.DataFrame

Parameters#

name

type

default

constraint

description

panel

pd.DataFrame

—

—

Input panel. Each column is a variable; rows are time periods. Series is promoted to a single-column DataFrame internally.

tcode_map

dict[str, int]

—

—

Mapping from column name to McCracken-Ng t-code integer 1..7. Columns not in tcode_map are passed through unchanged.

Returns#

pd.DataFrame — scalar result.

Behavior#

Each FRED-MD/QD series ships with a transformation code (1-7) mapping to a stationarity transform. apply_official_tcode runs the canonical mapping per series:

  • 1 = level

  • 2 = first difference

  • 3 = second difference

  • 4 = log

  • 5 = first difference of log (≈ growth rate)

  • 6 = second difference of log

  • 7 = log diff of (1 + growth rate)

Applied per-origin within walk-forward to avoid look-ahead.

When to use

Default for FRED-based studies. Canonical replication path.

In recipe context#

Set params.transform_policy = "apply_official_tcode" in the relevant layer to activate this op within a recipe:

# Layer L2 recipe fragment
params:
  transform_policy: apply_official_tcode

References#

  • macroforecast design Part 2, L2: ‘preprocessing is the only layer with a strict A→B→C→D→E execution order; every cell follows the same pipeline.’

  • McCracken & Ng (2016) ‘FRED-MD: A Monthly Database for Macroeconomic Research’, JBES 34(4). (doi:10.1080/07350015.2015.1086655)