TechnicalAnalysis.Net 3.0.0

TechnicalAnalysis.NET

TechnicalAnalysis.NET provides a C# API for technical indicators calculated by TA-Lib. It accepts price arrays or candle data and can produce a signal using either the indicator's default logic or your own.

This branch contains the next major version, 3.0.0, with breaking API and data ordering changes. See Migrating to 3.0 before upgrading.

Requirements and installation

The main TechnicalAnalysis.Net package targets .NET 10. Its separate TechnicalAnalysis.Net.Contracts project targets .NET Standard 2.0.

Once version 3.0.0 is published to NuGet, install it with:

dotnet add package TechnicalAnalysis.Net --version 3.0.0

Until then, build and reference the projects in this repository to use the version described here. Merely merging this branch does not publish a package.

Data ordering and results

• Provide input arrays and candle lists in newest-first order: index 0 is the latest observation, index 1 is the previous observation, and so on. For indicators with multiple input arrays, the arrays must have the same length and values at the same index must describe the same observation.
• On a successful calculation, each output series has the same length and ordering as the input. Positions without enough history are double.NaN at the end of the series. Inputs are not reordered in place.
• Success reports whether the calculation completed. A valid request with insufficient nonempty history can succeed with no valid output yet. On a failed calculation, inspect Message before accessing output series.
• series.HasValidData() checks that index 0 exists and contains a finite number. series.CountValidData() counts finite values in the whole series; it does not guarantee that they are consecutive. These extension methods require using TANet.Util.Extensions;.
• Check an older value with double.IsFinite(series[index]) after checking that the index is within the array. A valid value at index 0 does not imply that every older value is valid.

Example: simple moving average

using System;
using TANet.Core;
using TANet.Util.Extensions;

double[] close = [15, 14, 13, 12, 11]; // newest first
var result = Indicators.Sma(close, period: 3);

if (!result.Success)
{
    Console.WriteLine(result.Message);
}
else if (result.Ma.HasValidData())
{
    Console.WriteLine($"Latest SMA: {result.Ma[0]}"); // 14
    Console.WriteLine($"Signal: {result.IndicatorSignal}");
    Console.WriteLine($"Finite values: {result.Ma.CountValidData()}"); // 3

    if (result.Ma.Length > 1 && double.IsFinite(result.Ma[1]))
        Console.WriteLine($"Previous SMA: {result.Ma[1]}"); // 13
}
else
{
    Console.WriteLine("No finite value for the latest observation yet.");
}

The resulting series is [14, 13, 12, NaN, NaN] in newest-first order.

Custom signals

Many indicators accept a signalLogic callback. For example, SMA receives the original input and the output series, both newest first:

using TANet.Contracts.Enums;
using TANet.Core;

double[] close = [15, 14, 13, 12, 11];
var result = Indicators.Sma(close, 3, (input, output) =>
    input[0] > output[0] ? IndicatorSignal.Buy :
    input[0] < output[0] ? IndicatorSignal.Sell :
    IndicatorSignal.Stay);

The callback is invoked only when the output needed for the current signal is finite. With insufficient history, the result can still succeed, but the signal is Stay and the callback is not called. Default signals vary by indicator; ATR and Bollinger Bands return Stay unless custom logic is provided. Signals are simple rules, not a substitute for an evaluated trading strategy.

Candle input

Indicators that need multiple price series can also accept a List<Candle>:

using System.Collections.Generic;
using TANet.Contracts.Models;
using TANet.Core;

var candles = new List<Candle> // newest first
{
    new() { High = 14, Low = 12, Close = 13 },
    new() { High = 15, Low = 13, Close = 14 },
    new() { High = 12, Low = 10, Close = 11 },
    new() { High = 13, Low = 11, Close = 12 },
    new() { High = 11, Low = 9, Close = 10 }
};

var result = Indicators.Cci(candles, period: 3);

Check result.Success and the relevant output series as in the SMA example.

Implemented indicators

• Aroon
• Aroon Oscillator
• ATR
• Bollinger Bands
• CCI
• EMA
• Extended MACD
• KAMA
• MACD
• MFI
• Parabolic SAR
• RSI
• SMA
• Stochastic
• TEMA
• Williams %R
• WMA

• AC — Accelerator/Decelerator Oscillator
• ACCBANDS — Acceleration Bands
• AD — Chaikin A/D Line
• ADOSC — Chaikin A/D Oscillator
• ADR — Average Day Range
• ADX — Average Directional Movement Index
• ADXR — Average Directional Movement Index Rating
• ALMA — Arnaud Legoux Moving Average
• AO — Awesome Oscillator
• APO — Absolute Price Oscillator
• ASI — Wilder Accumulative Swing Index
• AVGDEV — Average Deviation
• AVGPRICE — Average Price
• BBW — Bollinger BandWidth
• BETA — Beta
• BOP — Balance Of Power
• CE — Chandelier Exit
• CG — Center of Gravity Oscillator
• CHOP — Choppiness Index
• CHOPTR — Choppiness Index (True Range Box)
• CKSP — Chande Kroll Stop
• CMF — Chaikin Money Flow
• CMO — Chande Momentum Oscillator
• CMOU — Chande Momentum Oscillator (Unsmoothed)
• COPPOCK — Coppock Curve
• CORREL — Pearson's Correlation Coefficient (r)
• CRSI — Connors Relative Strength Index
• CTI — Correlation Trend Indicator
• CVI — Chaikin's Volatility
• DEMA — Double Exponential Moving Average
• DONCHIAN — Donchian Channels
• DOSC — Derivative Oscillator
• DPO — Detrended Price Oscillator
• DX — Directional Movement Index
• EFI — Elder's Force Index
• EMV — Arms Ease of Movement
• ER — Kaufman Efficiency Ratio
• ERI — Elder Ray Index (Bull Power / Bear Power)
• FISHER — Fisher Transform
• FOSC — Forecast Oscillator
• FRACTAL — Williams Fractal
• FRAMA — Fractal Adaptive Moving Average
• HA — Heikin-Ashi Candles
• HMA — Hull Moving Average
• HT_DCPERIOD — Hilbert Transform - Dominant Cycle Period
• HT_DCPHASE — Hilbert Transform - Dominant Cycle Phase
• HT_PHASOR — Hilbert Transform - Phasor Components
• HT_SINE — Hilbert Transform - SineWave
• HT_TRENDLINE — Hilbert Transform - Instantaneous Trendline
• HT_TRENDMODE — Hilbert Transform - Trend vs Cycle Mode
• IBS — Internal Bar Strength
• ICHIMOKU — Ichimoku Kinko Hyo
• IMI — Intraday Momentum Index
• KC — Keltner Channels
• KDJ — KDJ Stochastic
• KST — Know Sure Thing (Pring)
• KSTEXT — Know Sure Thing with controllable MA type
• KURTOSIS — Rolling Excess Kurtosis
• KVO — Klinger Volume Oscillator
• LINEARREG — Linear Regression
• LINEARREG_ANGLE — Linear Regression Angle
• LINEARREG_INTERCEPT — Linear Regression Intercept
• LINEARREG_SLOPE — Linear Regression Slope
• MA — Moving average (partially supported)
• MACDFIX — Moving Average Convergence/Divergence Fix 12/26
• MACDV — MACD-V, Volatility Normalised Momentum
• MAMA — MESA Adaptive Moving Average
• MARKETFI — Market Facilitation Index
• MASSI — Mass Index
• MAVP — Moving average with variable period
• MCGD — McGinley Dynamic
• MEDIAN — Rolling Median
• MEDPRICE — Median Price
• MIDPOINT — MidPoint over period
• MIDPRICE — Midpoint Price over period
• MINUS_DI — Minus Directional Indicator
• MINUS_DM — Minus Directional Movement
• MOM — Momentum
• NATR — Normalized Average True Range
• NVI — Negative Volume Index
• OBV — On Balance Volume
• PERCENTB — Bollinger Bands %B
• PERCENTILE — Percentile (nearest rank)
• PERCENTRANK — Percent Rank
• PLUS_DI — Plus Directional Indicator
• PLUS_DM — Plus Directional Movement
• PPO — Percentage Price Oscillator
• PSO — Premier Stochastic Oscillator
• PVI — Positive Volume Index
• PVO — Percentage Volume Oscillator
• PVT — Price Volume Trend
• QSTICK — Qstick
• RMA — Wilder's Smoothed Moving Average
• ROC — Rate of change : ((price/prevPrice)-1)*100
• ROCP — Rate of change Percentage: (price-prevPrice)/prevPrice
• ROCR — Rate of change ratio: (price/prevPrice)
• ROCR100 — Rate of change ratio 100 scale: (price/prevPrice)*100
• ROGERSSATCHELL — Rogers-Satchell Volatility
• RVI — Relative Volatility Index
• RVIR — Relative Volatility Index, refined high/low form
• RVOL — Relative Volume
• SAREXT — Parabolic SAR - Extended
• SI — Wilder Swing Index
• SKEW — Rolling Skewness
• SMI — Stochastic Momentum Index
• SQZMOM — Squeeze Momentum and Level
• STC — Schaff Trend Cycle
• STDDEV — Standard Deviation
• STOCHF — Stochastic Fast
• STOCHRSI — Stochastic Relative Strength Index
• SUPERTREND — SuperTrend
• SWAK_2PHP — Swiss Army Knife - Two-Pole High-Pass Filter
• SWAK_BP — Swiss Army Knife - Band-Pass Filter
• SWAK_BUTTER — Swiss Army Knife - Butterworth Filter
• SWAK_GAUSS — Swiss Army Knife - Gaussian Filter
• SWAK_HP — Swiss Army Knife - High-Pass Filter
• T3 — Triple Exponential Moving Average (T3)
• TRANGE — True Range
• TRIMA — Triangular Moving Average
• TRIX — 1-day Rate-Of-Change (ROC) of a Triple Smooth EMA
• TSF — Time Series Forecast
• TSI — True Strength Index
• TYPPRICE — Typical Price
• ULTOSC — Ultimate Oscillator
• VAR — Variance
• VHF — Vertical Horizontal Filter
• VIDYA — Variable Index Dynamic Average
• VORTEX — Vortex Indicator
• VWAP — Volume Weighted Average Price
• VWMA — Volume Weighted Moving Average
• WAD — Williams' Accumulation/Distribution
• WAVETREND — WaveTrend Oscillator
• WCLPRICE — Weighted Close Price
• ZIGZAG — Zig Zag
• ZLEMA — Zero-Lag Exponential Moving Average

The list above follows the TA-Lib function reference. MA is partially supported through the existing moving average methods; not all reference MA types are exposed. CE and MACDV appear in the current reference but not in the installed TA-Lib C# 0.8.3 API, so they require a dependency review.

The array-based Williams %R overload currently requires a volume parameter for compatibility, although the calculation does not use it. TA-Lib's other technical indicators are planned; candlestick patterns and mathematical functions are outside the current scope.

Migrating to 3.0

• Use double and double[] for price data and indicator outputs instead of decimal and decimal[].
• Supply observations newest first and read the latest output at index 0. Outputs retain the input length, with NaN for unavailable history.
• Check Success separately from HasValidData(): a successful calculation does not necessarily mean that the latest value is available yet. Update custom signal logic to use the newest-first series and to check any older values it accesses.
• The main package now targets .NET 10 and uses the official C# TA-Lib package. The legacy Cryptowatcher.TA-LIb-Core dependency and the public MovingAverageTypes.FromTaLib and MovingAverageTypes.ToTaLib methods that used its types have been removed. ToOfficialTaLib converts TANet's supported moving average types to the official TALib.MAType.

Build and test

Install the .NET 10 SDK, then run from the repository root:

dotnet restore TANet/TANet.sln
dotnet test TANet/TANet.sln --configuration Release

Contributing

Issues and pull requests are welcome. Describe the problem or proposed change, and include tests for behavioral changes. For indicator additions, follow the newest-first and warm-up contract above and compare calculations against the official TA-Lib implementation. Run the test suite before submitting a pull request.

License

See LICENSE.