# \[ANN\] OnlineResamplers.jl - High-Performance Real-Time Financial Data Resampling
**URL:** https://discourse.julialang.org/t/ann-onlineresamplers-jl-high-performance-real-time-financial-data-resampling/132653
**Category:** Package Announcements
**Created:** [September 25, 2025, 12:27pm UTC](https://discourse.julialang.org/t/ann-onlineresamplers-jl-high-performance-real-time-financial-data-resampling/132653 "2025-09-25T12:27:23Z")
**Posts on this page:** 2
**Page:** 1
### Author:  [@FemtoTrader](https://discourse.julialang.org/u/FemtoTrader)
#### Post date: [September 25, 2025, 12:27pm UTC](https://discourse.julialang.org/t/ann-onlineresamplers-jl-high-performance-real-time-financial-data-resampling/132653/1 "2025-09-25T12:27:24Z")
Hi,
Iβm excited to announce **[OnlineResamplers.jl](https://femtotrader.github.io/OnlineResamplers.jl/dev/)**, a high-performance Julia package for real-time resampling of financial market data! π
## What is [OnlineResamplers.jl](https://github.com/femtotrader/OnlineResamplers.jl)?
OnlineResamplers.jl provides efficient streaming algorithms for aggregating tick-level market data into OHLC candlesticks and other time-based formats. Built on top of [OnlineStatsBase.jl](https://github.com/joshday/OnlineStatsBase.jl), it offers constant memory usage and zero-allocation operations for processing financial data streams.
## Key Features
π **Real-time Processing** : Stream market data with constant memory usage - no need to store historical data in memory
π **Multiple Resampling Methods** :
- OHLC (Open, High, Low, Close) candlesticks
- Mean price aggregation
- Volume sum aggregation
π’ **Parametric Types** : Full support for custom numeric types like `FixedPointDecimals.jl` for precise financial calculations
β‘ **High Performance** : Type-stable operations with zero allocations during steady-state processing
π **Parallel Processing** : Built-in merge operations for distributed computing scenarios
𧩠**OnlineStatsBase Integration** : Seamless compatibility with Juliaβs online statistics ecosystem
## Quick Example
```julia
using OnlineResamplers OnlineStatsBase, Dates
# Create a 1-minute OHLC resampler
resampler = MarketResampler(Minute(1))
# Process streaming market data
data = MarketDataPoint(DateTime(2024, 1, 1, 9, 30, 0), 100.0, 1000.0)
fit!(resampler, data)
# Get results
result = value(resampler)
println("OHLC: $(result.price.ohlc)") # OHLC(100.0, 100.0, 100.0, 100.0)
println("Volume: $(result.volume)") # 1000.0
```
## Why OnlineResamplers.jl?
Traditional approaches to market data resampling often require loading entire datasets into memory or using complex windowing mechanisms. OnlineResamplers.jl takes a different approach by:
1. **Processing data point-by-point** as it arrives
2. **Automatically handling time window transitions**
3. **Maintaining constant memory usage** regardless of data volume
4. **Supporting high-precision numeric types** for financial applications
This makes it ideal for:
- Real-time trading systems
- Market data processing pipelines
- Financial analysis applications
- Any scenario requiring efficient time-series aggregation
## Advanced Usage
The package supports sophisticated scenarios like custom numeric types:
```julia
using FixedPointDecimals
# High-precision resampler
resampler = MarketResampler{DateTime, FixedDecimal{Int64,4}, FixedDecimal{Int64,2}}(
Minute(1), price_method=:ohlc
)
# Process high-precision data
data = MarketDataPoint{DateTime, FixedDecimal{Int64,4}, FixedDecimal{Int64,2}}(
DateTime(2024, 1, 1, 9, 30, 0),
FixedDecimal{Int64,4}(100.5012),
FixedDecimal{Int64,2}(1000.50)
)
```
## Installation
```julia
using Pkg
Pkg.add(url="https://github.com/femtotrader/OnlineResampler.jl")
```
## Documentation & Examples
- π **Documentation** : Available in the [docs/](https://github.com/femtotrader/OnlineResampler.jl/tree/main/docs) directory
- π§ **API Reference** : Comprehensive function documentation
- π― **Tutorial** : Step-by-step guide from basic to advanced usage
- π‘ **Examples** : Real-world usage scenarios
## Performance
OnlineResampler.jl is designed for high-frequency data processing. The streaming approach means:
- **O(1) memory usage** per resampler instance
- **Zero allocations** during steady-state processing
- **Type-stable operations** for maximum performance
- **Parallel processing support** via merge operations
## Get Involved
The package is open source and welcomes contributions! Whether youβre working with financial data, time-series analysis, or online statistics, Iβd love to hear about your use cases and feedback.
**Repository** : [GitHub - femtotrader/OnlineResamplers.jl: high-performance Julia package for real-time resampling of financial market data](https://github.com/femtotrader/OnlineResampler.jl)
**Issues** : [GitHub Β· Where software is built](https://github.com/femtotrader/OnlineResampler.jl/issues)
* * *
_Built with β€ for the Julia community_
Would love to hear your thoughts and feedback! Has anyone been working on similar streaming data aggregation problems?
---
### Author:  [@Snowy](https://discourse.julialang.org/u/Snowy)
#### Post date: [September 25, 2025, 1:57pm UTC](https://discourse.julialang.org/t/ann-onlineresamplers-jl-high-performance-real-time-financial-data-resampling/132653/2 "2025-09-25T13:57:10Z")
Congrats on the package release and really great job on the documentation. Tons of tutorials, guides, edge cases, etc. We need more packages to have this robust documentation!