Esc

    The API is under testing and may change before release.

    Helixio 0.0

    Helixio is a quantitative finance library that prices fixed-income trades and measures their risk. The engine is the Rust crate helixrs, and the Python package helixio reaches the same engine, so both return the same values bit for bit. This page is the front door of the 0.0 documentation line. It tells you how the pages are organised and how to read one.

    What you need

    You need an installed build of the library; see Install helixio and check it runs. You need nothing else to read the pages: they are plain Markdown.

    Every page in this line describes one pinned revision of the 0.0.x series. The pin is recorded in the line's pin.toml, and every example on every page ran against it.

    The steps

    1. Start with Getting started. It installs the library, prices a first trade end to end, and explains the ideas the rest of the tree assumes.

    2. Read The market. Every later page builds its market the way it teaches: from a desk's quote sheet, in one call.

    3. Pick the section that matches your question. Each section answers one kind of question.

    Section Read it for
    Getting started installing, a first trade, the picture every page builds on, and coding agents
    Foundations dates, tenors, calendars, day counts, schedules, indices, templates, meeting schedules, listed futures, errors and vocabularies
    The market quote sheets and recipes, curve helpers, the dated market, data sources, books that carry their market
    Curves what each kind of curve stores, how it interpolates, how you read it, and where it came from
    Calibration solving curves and models from quotes: bootstrap, joint solves, credit, inflation, cross-currency, Hull-White, SABR, the Jacobian
    Cashflows legs and coupons, CMS and inflation coupons, coupon options, cashflow records
    Products one page per instrument, grouped into rates, bonds, credit, FX and financing
    Construction templates and product builders, counterparties and CSAs, the template registry
    How-to guides step-by-step recipes for desk tasks: price a swap from quotes, strike at the fair rate, value a cross-currency swap by leg
    Markets the conventions each market applies, and the rate and FX vol surfaces

    This line is being rewritten a section at a time. The sections still to come are risk, and scenarios and P&L.

    4. Use the API reference when you know the name. The reference lists every name both surfaces export. The sections above explain what those names compute.

    5. Know which kind of page you're on. Each page does one job, and its headings tell you which:

    • A tutorial teaches by doing: Before you start, Steps, What you built, Next steps.
    • A how-to guide gets a task done: Before you start, Steps, Check it worked, Related.
    • A reference page is for looking things up: Quick example, How it's priced, API, Defaults and limits, See also.
    • An explanation page says why things work the way they do, and ends with See also.

    Pages not yet rewritten keep the older six-part shape: a framing paragraph, Market context, Theory, From formula to API (each symbol mapped to the Rust and the Python call), Examples, and See also.

    What to check

    You can trust a code block or a number on these pages, because neither is typed by hand:

    • Every code block is copied from a file that runs against the pinned release.
    • Every output block is the text that file printed.
    • Every example runs in Rust and in Python, and the two must print the same text.
    • Every example asserts its result against the formula the page states.

    Prose is not checked the same way. If a sentence and an output block disagree, believe the output block.

    On rewritten pages, every code block is a whole program you can copy and run. Its checks run too, but they aren't shown. Some explanation pages, and pages in the older shape, also show a by hand version that rebuilds the same result from lower-level calls, so you can see what the short way assembles.

    Where next

    • To install and price something, go to Getting started.
    • To write helixio code with Claude Code or Codex, install the skill from Coding agents.
    • To build a market from your own quotes or data, go to The market.
    • To look up a name, go to the API reference.

    Credits

    Helixio owes its shape to three open-source projects that came first. The library was tested against them and developed with them open as references. It depends on none of them at build or run time, and neither do these pages.

    Project How Helixio uses it
    QuantLib The reference for conformance testing: the library prices each fixture in QuantLib, and its tests must reproduce the result or record why they differ. The holiday tables are generated from it.
    Open Source Risk Engine (ORE) A reference while developing scenarios and cross-currency swaps.
    Strata A reference while developing schedules, index observations, stubs and compounding.

    Thanks also go to Luigi Ballabio, who maintains QuantLib and wrote A QuantLib Guide, which was read while developing and testing Helixio.

    From the 0.0 line at revision bf7f2e2. Every printed number is that revision's own output.