Visualization and reports

AbaQuant separates numerical workflows from presentation workflows. Model, portfolio, credit, market-data, calibration, and dashboard objects can create visualizations and exportable reports.

Visualization setup

from abaquant.visualization import (
    VisualizationTheme,
    configure_visualization,
    visualization_theme,
    save_figure,
)

theme = VisualizationTheme(title="Research", backend="matplotlib")
configure_visualization(theme)

Use a temporary theme in a context manager:

with visualization_theme(VisualizationTheme(backend="plotly")):
    fig = model.visualize(chart="price_surface", option_type="call")

Common chart entry points

Object or namespace	Typical charts
Option model	price curves, Greek curves, price surfaces, Greek surfaces, intrinsic/extrinsic decomposition.
Option strategy	payoff profile, component payoff, net profit.
Option-chain analytics	IV smile, IV surface, skew, term structure, rich/cheap table, open-interest heatmap.
Portfolio allocator	weights, cumulative returns, correlation, risk contribution.
Portfolio backtest	equity curve, drawdown, rolling metrics, calendar returns, contribution, trades.
Credit assessment	metric dashboard, score/band visualization, scenario chart.
Market ticker/universe	price history, financial-statement charts, universe performance.
Calibration result	model-versus-market fit, residual diagnostics.
Risk dashboard	risk contribution, drawdown, correlation, credit score summary.

Save figures

fig = allocator.visualize(chart="correlation")
save_figure(fig, "portfolio_correlation.png")

Backends return different object types. Use save_figure() or the built-in filename arguments on high-level visualization methods when available.

Reports

Reports are built from reusable sections, metrics, and tables.

from abaquant.reports import ExportableReport, ReportSection, ReportTable

Most high-level objects expose .report():

option_report = model.report(option_type="call")
portfolio_report = allocator.report()
backtest_report = backtest.report()
credit_report = assessment.report()
dashboard_report = dashboard.report()

Export formats

written = option_report.save(
    "reports",
    "option_report",
    formats=("markdown", "html", "pdf"),
)

Format	Use case
Markdown	Plain-text review, Git diffs, notebooks, README fragments.
HTML	Browser review, richer tables, lightweight sharing.
PDF	Self-contained static report; generated by a lightweight built-in writer.

The PDF exporter is intentionally simple and pure Python. It is useful for compact static reports, not for complex print-layout publishing.

Report provenance

Reports can include generated metadata and provenance inherited from source objects. This is useful when a report combines live provider data, cached financial statements, rate curves, and model transformations.

Common pitfalls

• Creating many Matplotlib figures without closing them can trigger open-figure warnings.

• Plotly output requires the optional visualization dependencies.

• Figure object APIs differ by backend.

• Visualizations are explanatory aids; numerical result objects are the source of truth.