Kand - Blazingly Fast Technical Analysis Library

Documentation: Rust - Python | Repository: GitHub

Kand: Blazingly Fast Technical Analysis Library in Rust, Python

Why Kand?

🚀 Blazing Fast Built in Rust for elite performance and safety, rivaling top libraries like TALib.
🔥 GIL-Free Unlocks Python’s GIL for seamless multi-threading—unlike TALib’s single-threaded limits.
⏱️ O(1) Incremental Speed Lightning-fast incremental updates, outpacing traditional batch methods.
⚡️ Zero-Copy Native NumPy integration with zero-copy data passing—no overhead, just speed.
🛠️ One-Line Install Skip TALib’s complex C library setup—install with a single command.
💻 Cross-Platform Runs effortlessly on macOS, Linux, and Windows.

Discover more benefits in our comprehensive documentation.

Python API

The Python interface of kand leverages PyO3 for ultra-low latency bindings (~7ns overhead) to the Rust core, seamlessly integrating with NumPy for zero-copy operations and true thread-safe calculations. Below are examples for batch and incremental usage.

import numpy as np
from kand import ema

# Batch EMA computation with zero-copy NumPy integration
# Input: NumPy array of prices (float64)
# Output: Array of EMA values based on period and default smoothing factor
prices = np.array([10.0, 11.0, 12.0, 13.0, 14.0], dtype=np.float64)
ema_values = ema(prices, period=3)  # Uses default smoothing factor k=2/(period+1)

# Incremental EMA update for streaming data
# Input: New price and previous EMA; constant-time update
prev_ema = 13.5
new_price = 15.0
new_ema = ema_incremental(new_price, prev_ema, period=3)  # Default k=2/(period+1)

Key Features:

Zero-Copy: Operates directly on NumPy arrays, avoiding memory duplication.
GIL-Free: Rust backend releases the Python GIL, enabling parallel execution.
Incremental Updates: O(1) complexity for real-time applications.

Rust API

The Rust interface in kand provides a high-performance, type-safe implementation of EMA with flexible parameter control. It supports both Vec and ndarray inputs for batch and incremental calculations, as shown below.

use kand::ohlcv::ema;
use ndarray::Array1;

// Batch EMA calculation over a price series
// Input: Price vector, period, optional smoothing factor (None for default k=2/(period+1))
// Output: Writes EMA values to a pre-allocated buffer
let prices = vec![10.0, 11.0, 12.0, 13.0, 14.0];
let mut ema_values = vec![0.0; prices.len()];
ema::ema(&prices, 3, None, &mut ema_values)?;  // Default k=2/(4)=0.5

// Batch EMA with ndarray for scientific workflows
let prices = Array1::from_vec(vec![10.0, 11.0, 12.0, 13.0, 14.0]);
let mut ema_values = Array1::zeros(prices.len());
ema::ema(&prices, 3, None, &mut ema_values)?;  // Default k=2/(4)=0.5

// Constant-time incremental EMA update
// Input: New price, previous EMA, period, optional smoothing factor
let prev_ema = 13.5;
let new_price = 15.0;
let new_ema = ema::ema_incremental(new_price, prev_ema, 3, None)?;  // Default k=0.5

Key Features:

Memory Efficiency: Leverages mutable buffers (&mut Vec<f64> or &mut Array1<f64>) to store results, slashing memory allocations.
Error Handling: Returns Result<(), KandError> or Result<f64, KandError> for reliable failure detection (e.g., invalid period, NaN inputs).
Incremental Design: O(1) updates tailored for real-time systems.

Setup

Python

Get started with Kand in one command - no extra configuration needed:

pip install kand

Rust

You can take latest release from crates.io, or if you want to use the latest features / performance improvements point to the main branch of this repo.

[dependencies]
kand = { git = "https://github.com/rust-ta/kand", rev = "<optional git tag>" }

Recommend Rust version >=1.80.

Functions List

OHLCV Based

Statistical Analysis

Contributing

We are passionate about supporting contributors of all levels of experience and would love to see you get involved in the project. See the contributing guide to get started.

License

This project is licensed under either of the following licenses, at your option: - Apache License, Version 2.0, (LICENSE-APACHE or https://www.apache.org/licenses/LICENSE-2.0) - MIT license (LICENSE-MIT or https://opensource.org/licenses/MIT)

Unless you explicitly state otherwise, any contribution intentionally submitted for inclusion in kand by you, as defined in the Apache-2.0 license, shall be dually licensed as above, without any additional terms or conditions.