comparison_output_form
Back to L2.5 | Browse all axes | Browse all options
Axis
comparison_output_formon sub-layerL2_5_A_comparison_axis(layerl2_5).
Sub-layer
L2_5_A_comparison_axis
Axis metadata
Default:
'multi'Sweepable: False
Status: operational
Operational status summary
Operational: 5 option(s)
Future: 0 option(s)
Options
difference_table – operational
Table of (after - before) per metric per series.
L2.5.A comparison output form difference_table.
This option configures the comparison_output_form axis on the L2_5_A_comparison_axis sub-layer of L2.5; output is emitted under manifest.diagnostics/l2_5/L2_5_A_comparison_axis/ alongside the other selected views.
When to use
Quantifying the magnitude of cleaning shifts; ordered by absolute change for easy ranking.
References
macroforecast design Part 4: ‘diagnostic layers default-off; non-blocking; produce JSON + matplotlib views attached to manifest.diagnostics/.’
Related options: side_by_side_summary, overlay_timeseries, distribution_shift, multi
Last reviewed 2026-05-05 by macroforecast author.
distribution_shift – operational
Distribution-overlap statistics (KS / histogram).
L2.5.A comparison output form distribution_shift.
This option configures the comparison_output_form axis on the L2_5_A_comparison_axis sub-layer of L2.5; output is emitted under manifest.diagnostics/l2_5/L2_5_A_comparison_axis/ alongside the other selected views.
When to use
Detecting cleaning steps that materially change distributions, not just summary statistics.
References
macroforecast design Part 4: ‘diagnostic layers default-off; non-blocking; produce JSON + matplotlib views attached to manifest.diagnostics/.’
Related options: side_by_side_summary, difference_table, overlay_timeseries, multi
Last reviewed 2026-05-05 by macroforecast author.
multi – operational
Render every comparison output form together.
L2.5.A comparison output form multi.
This option configures the comparison_output_form axis on the L2_5_A_comparison_axis sub-layer of L2.5; output is emitted under manifest.diagnostics/l2_5/L2_5_A_comparison_axis/ alongside the other selected views.
When to use
Comprehensive cleaning audit; recommended default for first-time runs.
References
macroforecast design Part 4: ‘diagnostic layers default-off; non-blocking; produce JSON + matplotlib views attached to manifest.diagnostics/.’
Related options: side_by_side_summary, difference_table, overlay_timeseries, distribution_shift
Last reviewed 2026-05-05 by macroforecast author.
overlay_timeseries – operational
Time-series overlay of before / after for each series.
L2.5.A comparison output form overlay_timeseries.
This option configures the comparison_output_form axis on the L2_5_A_comparison_axis sub-layer of L2.5; output is emitted under manifest.diagnostics/l2_5/L2_5_A_comparison_axis/ alongside the other selected views.
When to use
Visual confirmation that cleaning preserves dynamics; spotting unexpected shape changes.
References
macroforecast design Part 4: ‘diagnostic layers default-off; non-blocking; produce JSON + matplotlib views attached to manifest.diagnostics/.’
Related options: side_by_side_summary, difference_table, distribution_shift, multi
Last reviewed 2026-05-05 by macroforecast author.
side_by_side_summary – operational
Side-by-side summary statistics for the chosen pair.
L2.5.A comparison output form side_by_side_summary.
This option configures the comparison_output_form axis on the L2_5_A_comparison_axis sub-layer of L2.5; output is emitted under manifest.diagnostics/l2_5/L2_5_A_comparison_axis/ alongside the other selected views.
When to use
Default tabular comparison; matches L1.5.B summary metrics applied to both pre and post.
References
macroforecast design Part 4: ‘diagnostic layers default-off; non-blocking; produce JSON + matplotlib views attached to manifest.diagnostics/.’
Related options: difference_table, overlay_timeseries, distribution_shift, multi
Last reviewed 2026-05-05 by macroforecast author.