Charts

A G-Cubed chartpack selects projection series and organises them into a report. Changing a chartpack changes presentation, not the model or its results.

Related YAML documentation: chartpack YAML, series selectors, labels and styles and setup-generated chartpacks.

A small chartpack

- chartpack:
    format: gcubed-chartpack
    version: 1
    title: Policy experiment
- chart:
    title: Inflation
    series:
      - variable: INFL(USA)
      - variable: INFL(CHN)

Build 199 chartpacks use .yaml and the chart-template-2.0.html reporting template.

Select model members

Use an exact variable for a known line or a selector to expand a prefix:

- chart:
    title: Real GDP by region
    series:
      - variable_prefix: GDPR
        selectors:
          regions: all

Selectors are resolved against the target projections. Every selected series in one chart must have the same units.

Generate chartpacks from setup

Setup uses a concise, model-aware declaration and writes standalone chartpack YAML. The two grammars differ: setup uses ordered selector records and the style key colour; a standalone chartpack uses selector mappings and color.

See generate chartpacks for single and multiple output files.

Projection CSV files

G-Cubed reports use charting projections or deviations written by a run script. These projection data files remain CSV. Their leading columns identify and describe each series:

Column	Meaning
first column	Full G-Cubed variable name
prefix	Variable-name prefix
set_member_descriptions	Descriptions of the selected set members
description	Description of the variable
units	Units of measurement
vector	G-Cubed vector type
region	G-Cubed region associated with the variable
YYYY columns	Projection values for each year, in order

When creating a chartpack, start from a build 199 YAML chartpack for the same model and adapt its charts and selectors. Lists of related variable names in a projection CSV can help identify useful series, but the chartpack itself must remain YAML.

Produce the report

The maintained baseline or experiment run script selects the standalone chartpack and produces the report. Setup does not run the model and a standalone chartpack does not calculate derived variables by itself.

This page does not currently provide a supported build 199 YAML chartpack browser application. Use the normal Python reporting workflow documented by the maintained run script.