Python

Download Python packages as folder ZIPs or wheels. Choose the package for your operating system, architecture and Python version. Folder ZIPs contain only QuantSupport; install dependencies separately with python -m pip install pandas.

The Python package provides native QuantSupport instruments, market construction, valuation and risk calculations. Define the contract and position, configure the market, then request values, cashflows or sensitivities from a pricing context. The user guide develops this workflow with complete examples; the studies apply it to larger quantitative finance calculations.

• Download Python packages
• Python user guide
  • A position, a market, and a valuation
  • Constructing and selecting a market
  • Valuation, quotation, and model choice
  • Risk, netting, and extensibility
  • Reading path
• Computational studies
  • Curve construction
  • Swaps and collateral currencies
  • Fixed income and loans
  • Rate options and volatility
  • Short-rate simulation
  • Portfolio risk and benchmarking
  • Scripted payoffs
  • Exposure and XVA
  • Reproduction
• API reference
  • Time
  • Market Data
  • Instruments
  • Pricing
  • Risk and XVA

Each class has its own declaration page with constructors, parameter details, methods, properties, and class attributes. Domain packages group related types; the package root re-exports the same class objects for compatibility. Each domain is divided into focused subsections, such as Calendars, Yield Curves, Fixed Income, and Collateral and Netting.

Conventions

Signatures and parameter descriptions document accepted runtime input forms. The reference is generated from the packaged stubs and their docstrings. A default displayed as ... comes from the native implementation; its parameter description states the actual default or how it is selected. Dates accept the documented Date, ISO text, or datetime.date inputs. Rates and spreads are decimal annual rates unless a parameter specifies other units.

Properties without setters are read-only. Context input properties with setters invalidate the dependent constructed market when reassigned. Method calls such as evaluate() perform operations rather than exposing attributes.

QuantSupport failures raise QuantSupportError or one of its subclasses. Native and Python interfaces share stable error codes and native variant names. Type, value, state, I/O, and unsupported-operation subclasses also inherit the corresponding Python built-in exception, preserving existing catches.

Exceptions

exception quantsupport.QuantSupportError

Shared structured failure for construction, input conversion, market-data resolution, and evaluation. Inspect code, native_type, message, operation, path, expected, actual, contexts, and details. Fields that do not apply are None. __cause__ preserves the original exception chain; to_dict() exports the same fields and nested causes.

The Python subclasses are QuantSupportTypeError, QuantSupportValueError, QuantSupportStateError, QuantSupportIOError, and QuantSupportNotImplementedError.

try:
    quotes = qs.QuoteStore.from_dict(market_inputs)
except qs.QuantSupportError as error:
    print(error.code, error.operation, error.path)
    diagnostics = error.to_dict()

For example, an invalid quote level can report code="deserialization", native_type="DeserializationErr", and path="quotes[1].mid". A missing quote instead reports code="not_found" and native_type="NotFoundErr". These come from native QSError categories, rather than message matching. File failures include the file path in details; JSON failures also retain line and column when the parser supplies them.

Python protocols retain normal behavior for attribute lookup, missing mapping keys, iteration termination, and indexing: AttributeError, KeyError, StopIteration, and IndexError continue to work with Python’s containers and inspection functions.

Installation, a first valuation, protocols, and reproducibility are explained locally in A position, a market, and a valuation. The binding README also contains source-build instructions and development checks.