task_charts__bar_3d__pairwise_comparison_count

Contract

1. Domain: charts
2. Scene id: bar_3d
3. Source implementation scene: charts/bar_3d
4. Query id: single
5. Semantic query details are recorded in query_id and trace params.

Implementation

1. Registered class: trace_tasks.tasks.charts.bar_3d.pairwise_comparison_count.ChartsThreeDBarPairwiseComparisonCountTask
2. Prompt lookup scene: charts/bar_3d
3. Generation is deterministic from instance_seed, explicit params, prompt bundle, renderer config, and code versions.
4. Answers and annotation are produced from the same metadata execution trace.

Annotation Contract

1. Answer schema: integer_count.
2. Annotation schema: segment_set.
3. Annotation is a segment_set; each segment is two [x, y] pixel points formatted [[x0, y0], [x1, y1]] and connects the top-center points of the two compared bars for one counted category.
4. Renderer context such as legends, axes, decorative labels, titles, and distractor text is metadata unless the task explicitly asks for it as annotation.

Program Contract

Program: count(category where compare(value(category,series_a), value(category,series_b), relation=greater_than)); output=integer_count; annotation=segment_set(comparison_bar_top_center_pairs); scene=bar_3d; scope=pairwise_comparison_count

Candidate set: the visible 3D bars grouped by category and series inside the pairwise_comparison_count objective scope. Operands: prompt-bound labels, categories, series names, thresholds, intervals, references, and encoded chart values, plus the task's prompt-bound target operands when present. Operation: evaluate count over the candidate set using the filters, comparisons, aggregations, rankings, projections, or counterfactual edits named in the program expression; generation enforces a unique final answer. Output binding: answer is the integer_count value bound by integer_count. Annotation witnesses: segment_set witnesses bound by segment_set(comparison_bar_top_center_pairs). Annotation is a segment_set; each segment is two [x, y] pixel points formatted [[x0, y0], [x1, y1]] and connects the top-center points of the two compared bars for one counted category. Renderer context such as legends, axes, decorative labels, titles, and distractor text is metadata unless the task explicitly asks for it as annotation. Query ids: single.

Reasoning Operations

Families: filtering, counting, comparison

Query Details

Query id	Program signature	Answer schema	Annotation schema
single	count.pairwise_series_comparison	integer_count	segment_set