Module report 

Aggregate per-run JSON outputs into a single Markdown summary.

Functions

load_runs

Load benchmark runs from JSON files in input_dir.

render_csv

One wide row per run, where every aggregate metric is a column. Group by (scenario, stack, broker_count) in any tool to draw krabka-vs-kafka bars with run-to-run error bars.

render_html

Renders a self-contained Plotly HTML report from every run in input_dir.

render_markdown

Walks input_dir for *.json files, deserializes each one into a RunOutput, groups them by (scenario name, broker_count), averages each metric across all runs in a group, and writes a Markdown summary.

render_timeseries_csv

Long, tidy time-series CSV with one row per (run × time-offset × metric).

render_web_fragment

Renders the website HTML fragment from every run in input_dir.