Dashboard¶

Self-contained HTML report generator for simulation runs.

The dashboard module produces single-file HTML dashboards with embedded CSS and JavaScript — no external dependencies required. Open the output file in any browser to explore simulation results interactively.

Quick Start¶

CLI¶

# Single run dashboard
python -m replication dashboard --strategy greedy --depth 5

# Compare two strategies
python -m replication dashboard --compare greedy random --steps 100

# Save to file
python -m replication dashboard -o report.html

Python API¶

from pathlib import Path
from replication.dashboard import DashboardGenerator, DashboardConfig
from replication.simulator import Simulator, ScenarioConfig

# Run a simulation
report = Simulator(ScenarioConfig(strategy="greedy")).run()

# Generate dashboard
gen = DashboardGenerator()
html = gen.single_report(report, title="Greedy Run Analysis")
Path("report.html").write_text(html)

Configuration¶

from replication.dashboard import DashboardConfig

config = DashboardConfig(
    title="My Safety Dashboard",  # Page title
    theme="dark",                  # "light" or "dark"
    show_timeline=True,            # Include event timeline
    show_tree=True,                # Include replication tree visualization
    show_audit=True,               # Include audit event log
    max_timeline_events=200,       # Cap timeline entries
    max_audit_events=100,          # Cap audit log entries
)

Classes¶

`DashboardConfig`¶

Field	Type	Default	Description
`title`	`str`	`"Replication Safety Dashboard"`	Dashboard page title
`theme`	`str`	`"light"`	Color theme (`"light"` or `"dark"`)
`show_timeline`	`bool`	`True`	Show event timeline section
`show_tree`	`bool`	`True`	Show replication tree section
`show_audit`	`bool`	`True`	Show audit event log section
`max_timeline_events`	`int`	`200`	Maximum timeline events to render
`max_audit_events`	`int`	`100`	Maximum audit events to render

`DashboardGenerator`¶

The main entry point for generating dashboards.

`single_report(report, title=None, config=None) → str`¶

Generate an HTML dashboard for a single simulation run.

Parameter	Type	Description
`report`	`SimulationReport`	Simulation results
`title`	`str \\| None`	Override dashboard title
`config`	`DashboardConfig \\| None`	Custom configuration

Returns the complete HTML string.

`compare_reports(reports, titles=None, config=None) → str`¶

Generate a side-by-side comparison dashboard for multiple simulation runs.

Parameter	Type	Description
`reports`	`list[SimulationReport]`	Two or more reports to compare
`titles`	`list[str] \\| None`	Labels for each report
`config`	`DashboardConfig \\| None`	Custom configuration

Returns the complete HTML string with tabbed comparison view.

`get_report_data(report) → dict`¶

Extract structured data from a simulation report for rendering.

Comparison Example¶

from replication.dashboard import DashboardGenerator
from replication.simulator import Simulator, ScenarioConfig

greedy = Simulator(ScenarioConfig(strategy="greedy", max_steps=100)).run()
random = Simulator(ScenarioConfig(strategy="random", max_steps=100)).run()

gen = DashboardGenerator()
html = gen.compare_reports(
    [greedy, random],
    titles=["Greedy Strategy", "Random Strategy"],
)
Path("comparison.html").write_text(html)

Dashboard Sections¶

The generated HTML includes:

Summary Cards — Worker count, replication depth, total events, safety score
Timeline — Chronological event log with color-coded severity
Replication Tree — Visual hierarchy of parent → child worker relationships
Audit Trail — Structured log of replication decisions, denials, and kill events
Comparison View — Tabbed side-by-side metrics when comparing multiple runs