Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/Nonanti/mathcore/llms.txt

Use this file to discover all available pages before exploring further.

This page covers everything you need to add MathCore to a Rust project: the two installation methods, a full breakdown of the available feature flags (including no_std configuration), the underlying dependencies, and a quick verification step to confirm your setup is working correctly.

​
Requirements

  • Rust 1.65 or later (MathCore uses the 2021 edition)
  • Cargo (ships with the standard Rust toolchain)

​
Installation Methods

cargo add mathcore
Both methods add MathCore with its default features enabled (std, parallel, and fft). See the Feature Flags section below to customise the build.

​
Feature Flags

MathCore’s feature system lets you pick exactly the capabilities you need, from a minimal no_std core up to the full-featured default build.
FeatureDefaultDescription
std✅Enables standard-library support: std::collections::HashMap, std-backed error types via thiserror, and nom/nalgebra std integration.
parallel✅Rayon-based data parallelism for matrix and batch operations. Requires std.
fft✅Fast Fourier Transform support via rustfft. Requires std.
full❌Alias that explicitly enables std, parallel, and fft together.
The default feature set is ["std", "parallel", "fft"]. In a standard Rust project targeting a desktop or server environment, you can omit the features key entirely and get all three.

​
Choosing a configuration

# No std, no parallelism, no FFT.
# Safe for embedded and bare-metal targets.
mathcore = { version = "0.3.1", default-features = false }

​
no_std usage

MathCore’s root crate is declared #![no_std] and depends only on alloc at its core, so it can target embedded systems and other bare-metal environments. When std is disabled:
  • HashMap<K, V> is backed by alloc::collections::BTreeMap instead of std::collections::HashMap.
  • thiserror, nom, and nalgebra all fall back to their no_std/alloc modes.
  • rayon and rustfft are not available (both require std).
[dependencies]
mathcore = { version = "0.3.1", default-features = false }
Disabling the std feature also disables parallel and fft automatically, because both declare std as a dependency in their feature definitions. You do not need to list them explicitly.

​
Dependencies

MathCore pulls in the following crates. All are widely used, well-maintained libraries with their own no_std support paths:
CrateVersionRole
nom7.1Expression parser combinators
num-complex0.4Complex number arithmetic
num-traits0.2Numeric trait abstractions
num-bigint0.4Arbitrary-precision integers
num-rational0.4Exact rational arithmetic (backed by num-bigint)
nalgebra0.32Matrix and linear algebra operations
thiserror2.0Ergonomic error-type derivation
rayon1.7Data parallelism (optional, parallel feature)
rustfft6.1Fast Fourier Transform (optional, fft feature)
nalgebra uses libm for floating-point math in no_std builds, so it does not require a system math library.

​
Verifying Your Installation

After adding MathCore, confirm everything compiles and the test suite passes: 
cargo test
You should see output ending with: 
test result: ok. X passed; 0 failed; 0 ignored
To run the bundled demo that exercises arithmetic, differentiation, integration, equation solving, and ASCII plotting, use: 
cargo run --example demo
To build optimised release artefacts (uses LTO and opt-level = 3 as configured in Cargo.toml): 
cargo build --release
To browse the API documentation locally, run cargo doc --open. This builds and opens the full rustdoc output in your browser without needing an internet connection.

Build docs developers (and LLMs) love

Get started for freeTalk to us