Nexus Engine

The Python client for Nexus Engine — a Rust-based calculation engine for financial data processing.

This package provides a fluent, chainable DSL for building data transformation and calculation pipelines. Define columns from SQL data sources, apply filters, joins, groupings, conditional logic, and hierarchical breakdowns — then send the pipeline to Nexus Engine for execution.

Note: Executing pipelines requires access to a running Nexus Engine instance and valid credentials. Contact your administrator or visit nexus-engine.io for access.

Installation

pip install nexus-engine[all]

nexusctl login

Quick Start

import nexus as nx

# Define source columns from a SQL data view
table = nx.TableBuilder([
    nx.sql_column("Category", nx.DataType.String, "positions", "asset_category"),
    nx.sql_column("Currency", nx.DataType.String, "positions", "local_currency"),
    nx.sql_column("MarketValue", nx.DataType.Float, "positions", "market_value"),
])

# Build a transformation pipeline
pipeline = (
    table
    .filter(nx.col("MarketValue") > nx.lit(10000))
    .group_by(
        by=["Category"],
        aggregations=[("MarketValue", "sum")]
    )
)

# Execute against a Nexus Engine server
result = pipeline.execute(
    params={"account_id": "ACC001", "date": "2025-01-31"})
print(result)

Type Support

The package includes full type stubs (.pyi) for IDE autocompletion and type checking with mypy or pyright.

Documentation

Full documentation and tutorials are available at nexus-engine.io.

License

Dual-licensed under MIT or Apache-2.0.