`sample_end_rule`#

Back to L1 | Browse all axes | Browse all options

Axis sample_end_rule on sub-layer l1_e (layer l1).

Sub-layer#

l1_e

Axis metadata#

Default: 'latest_available'
Sweepable: False
Status: operational

Operational status summary#

Operational: 2 option(s)
Future: 0 option(s)

Options#

`latest_available` – operational#

End at the panel’s last date.

Default. Uses the most recent observation in the bundled vintage.

Configures the sample_end_rule axis on l1_e (layer l1); the latest_available value is materialised in the recipe’s fixed_axes block under that sub-layer.

When to use

Default. Studies that want to use the full available history.

References

macroforecast design Part 1, L1: ‘data definition is the recipe layer that pins source, target, geography, and horizon – everything downstream branches off these choices.’

Related options: fixed_date

Last reviewed 2026-05-04 by macroforecast author.

`fixed_date` – operational#

Pin the end date in leaf_config (e.g., 2019-12-31).

Requires leaf_config.sample_end_date. Useful for pseudo-out-of-sample evaluation where the recipe wants to exclude the COVID window or stop at a paper’s reported sample.

When to use

Pre-COVID benchmark studies; matching a paper’s reported sample window.

References

macroforecast design Part 1, L1: ‘data definition is the recipe layer that pins source, target, geography, and horizon – everything downstream branches off these choices.’

Related options: latest_available

Parameters

name	type	default	constraint	description
`sample_end_date`	`str`	—	Required when sample_end_rule=fixed_date. ISO date string: full YYYY-MM-DD, partial YYYY-MM (normalized to last day of month), or YYYY (normalized to Dec 31). Must be >= sample_start_date when both are fixed.	Explicit panel end date. The L1 loader trims the panel so that no observation falls after this date.