README
¶
Ban Technical Analysis
banta is an event-based technical analysis framework. It updates and caches indicators on each K-line, and the indicator calculation results are reused globally. It aims to provide a high degree of freedom, high performance, simple and easy-to-use indicator framework.
Core Concept
Traditional technical analysis libraries such as ta-lib and pandas-ta are very popular and have undergone a lot of performance optimization. They are very fast when calculating hundreds or thousands of candlesticks at a time.
However, when using these libraries in your robot's real-time trading, you need to pass the data of the previous hundreds of candlesticks at the same time when you receive a new candlestick. If you are running more symbols or if you are running on 1m or even 1s, the calculation delay will be too large to be tolerated.
Many people have used TradingView. The PineScript it uses is an event-based technical analysis engine. It does not re-run the previous candlesticks when receiving a new candlestick, but uses the cached results from the previous candlesticks.
This is also the design concept of BanTA, based on events, running once for each candlestick, and using the cached results from the previous candlesticks.
How BanTA Works
In BanTA, Series sequence is the key type. Basically, all return values are sequences. The Series has the Data []float64 field, which records the values of the current indicator for each candlestick.
The most commonly used e.Close is the closing price sequence. e.Close.Get(0) is the current closing price of type float64;
Calculating the average is also simple: ma5 := ta.SMA(e.Close, 5), the returned ma5 is also a sequence;
Some indicators such as KDJ generally return two fields k and d kdjRes := ta.KDJ(e.High, e.Low, e.Close, 9, 3, 3).Cols, you can get an array of two sequences from Cols.
Supported indicators
Comparison of consistency with common indicator platform results
| banta | MyTT | TA-lib Class | TA-lib Metastock | Pandas-TA | TradingView |
|---|---|---|---|---|---|
| AvgPrice | ✔ | ✔ | ✔ | ✔ | ✔ |
| Sum | ✔ | ✔ | ✔ | ✔ | ✔ |
| SMA | T1 | ✔ | ✔ | ✔ | ✔ |
| EMA | T1 | T1 | ✔ | ✔ | T2 |
| EMABy1 | ✔ | T1 | T2 | T2 | T3 |
| RMA | -- | -- | -- | T1 | -- |
| VWMA | -- | -- | -- | ✔ | ✔ |
| WMA | ✔ | ✔ | ✔ | ✔ | ✔ |
| HMA | -- | -- | -- | ✔ | ✔ |
| TR | -- | ✔ | ✔ | ✔ | -- |
| ATR | T1 | ✔ | ✔ | T2 | T3 |
| MACD | T1 | T2 | T1 | ✔ | T3 |
| RSI | T1 | ✔ | ✔ | T2 | T3 |
| KDJ | T1 | T2 | T1 | T3 | ✔ |
| Stoch | -- | ✔ | ✔ | -- | ✔ |
| BBANDS | ✔ | ✔ | ✔ | ✔ | ✔ |
| Aroon | -- | ✔ | ✔ | ✔ | T1 |
| ADX | -- | ✔ | ✔ | T1 | T2 |
| ADXBy1 | -- | T1 | T1 | T2 | ✔ |
| PluMinDI | -- | ✔ | ✔ | -- | -- |
| PluMinDM | -- | ✔ | ✔ | -- | -- |
| ROC | ✔ | ✔ | ✔ | -- | ✔ |
| TNR/ER | -- | -- | -- | -- | -- |
| CCI | ✔ | ✔ | ✔ | ✔ | ✔ |
| CMF | -- | -- | -- | ✔ | ✔ |
| KAMA | -- | ✔~ | ✔~ | T1 | ✔~ |
| WillR | -- | ✔ | ✔ | ✔ | ✔ |
| StochRSI | -- | ✔ | ✔ | ✔ | ✔~ |
| MFI | ✔ | ✔ | ✔ | ✔ | ✔ |
| RMI | -- | -- | -- | -- | ✔~ |
| CTI | -- | -- | -- | ✔ | T1 |
| LinReg | -- | -- | -- | ✔ | ? |
| CMO | -- | ✔ | ✔ | ✔ | T1 |
| CMOBy1 | -- | T1 | T1 | T1 | ✔ |
| CHOP | -- | -- | -- | ✔ | T1 |
| ALMA | -- | -- | -- | ✔ | T1 |
| Stiffness | -- | -- | -- | -- | ✔ |
| PercentRank | -- | -- | -- | -- | ✔~ |
| CRSI | -- | -- | -- | -- | ✔~ |
| CRSIBy1 | -- | community | -- | -- | -- |
-- This platform does not have this indicator
✔ The calculation results are consistent with this platform
✔~ The calculation results are basically consistent with this platform (with some deviation)
Ti The calculation results are inconsistent with this platform
How to Use
package main
import (
"fmt"
ta "github.com/banbox/banta"
)
var envMap = make(map[string]*ta.BarEnv)
func OnBar(symbol string, timeframe string, bar *ta.Kline) {
envKey := fmt.Sprintf("%s_%s", symbol, timeframe)
e, ok := envMap[envKey]
if !ok {
e = &ta.BarEnv{
TimeFrame: timeframe,
BarNum: 1,
}
envMap[envKey] = e
}
e.OnBar(bar)
ma5 := ta.SMA(e.Close, 5)
ma30 := ta.SMA(e.Close, 30)
atr := ta.ATR(e.High, e.Low, e.Close, 14).Get(0)
xnum := ta.Cross(ma5, ma30)
if xnum == 1 {
// ma5 cross up ma30
curPrice := e.Close.Get(0) // or bar.Close
stopLoss := curPrice - atr
fmt.Printf("open long at %f, stoploss: %f", curPrice, stopLoss)
} else if xnum == -1 {
// ma5 cross down ma30
curPrice := e.Close.Get(0)
fmt.Printf("close long at %f", curPrice)
}
kdjRes := ta.KDJ(e.High, e.Low, e.Close, 9, 3, 3).Cols
k, d := kdjRes[0], kdjRes[1]
}
Documentation
¶
Index ¶
- Constants
- Variables
- func Cross(se *Series, obj2 interface{}) int
- func RemoveFromArr[T comparable](arr []T, it T, num int) []T
- type BarEnv
- type CCentre
- type CGraph
- func (c *CGraph) AddBar(e *BarEnv) *CGraph
- func (c *CGraph) AddBars(barId int, bars ...*Kline) *CGraph
- func (c *CGraph) AddPen(pen *CPen)
- func (c *CGraph) AddSeg(seg *CSeg)
- func (c *CGraph) Bar(v int) *Kline
- func (c *CGraph) Dump() []*DrawLine
- func (c *CGraph) NewPen(a, b *CPoint) *CPen
- func (c *CGraph) NewPoint(dirt float64, price float64, barId int) *CPoint
- func (c *CGraph) Parse()
- func (c *CGraph) Remove(o interface{}) bool
- type CPen
- type CPoint
- type CSeg
- type CTrend
- type CTwoPoint
- type CrossLog
- type DrawLine
- type Kline
- type Series
- func ADL(env *BarEnv) *Series
- func ADX(high *Series, low *Series, close *Series, period int) *Series
- func ADXBy(high *Series, low *Series, close *Series, period int, method int) *Series
- func ALMA(obj *Series, period int, sigma, distOff float64) *Series
- func ATR(high *Series, low *Series, close *Series, period int) *Series
- func Aroon(high *Series, low *Series, period int) *Series
- func AvgDev(obj *Series, period int) *Series
- func AvgPrice(e *BarEnv) *Series
- func BBANDS(obj *Series, period int, stdUp, stdDn float64) *Series
- func CCI(obj *Series, period int) *Series
- func CHOP(e *BarEnv, period int) *Series
- func CMF(env *BarEnv, period int) *Series
- func CMO(obj *Series, period int) *Series
- func CMOBy(obj *Series, period int, maType int) *Series
- func CRSI(obj *Series, period, upDn, roc int) *Series
- func CRSIBy(obj *Series, period, upDn, roc, vtype int) *Series
- func CTI(obj *Series, period int) *Series
- func ChaikinOsc(env *BarEnv, short int, long int) *Series
- func EMA(obj *Series, period int) *Series
- func EMABy(obj *Series, period int, initType int) *Series
- func ER(obj *Series, period int) *Series
- func HL2(h, l *Series) *Series
- func HLC3(h, l, c *Series) *Series
- func HMA(obj *Series, period int) *Series
- func HeikinAshi(e *BarEnv) *Series
- func Highest(obj *Series, period int) *Series
- func HighestBar(obj *Series, period int) *Series
- func KAMA(obj *Series, period int) *Series
- func KAMABy(obj *Series, period int, fast, slow int) *Series
- func KDJ(high *Series, low *Series, close *Series, period int, sm1 int, sm2 int) *Series
- func KDJBy(high *Series, low *Series, close *Series, period int, sm1 int, sm2 int, ...) *Series
- func LinReg(obj *Series, period int) *Series
- func LinRegAdv(obj *Series, period int, angle, intercept, degrees, r, slope, tsf bool) *Series
- func Lowest(obj *Series, period int) *Series
- func LowestBar(obj *Series, period int) *Series
- func MACD(obj *Series, fast int, slow int, smooth int) *Series
- func MACDBy(obj *Series, fast int, slow int, smooth int, initType int) *Series
- func MFI(e *BarEnv, period int) *Series
- func PercentRank(obj *Series, period int) *Series
- func PluMinDI(high *Series, low *Series, close *Series, period int) *Series
- func PluMinDM(high *Series, low *Series, close *Series, period int) *Series
- func RMA(obj *Series, period int) *Series
- func RMABy(obj *Series, period int, initType int, initVal float64) *Series
- func RMI(obj *Series, period int, montLen int) *Series
- func ROC(obj *Series, period int) *Series
- func RSI(obj *Series, period int) *Series
- func RSI50(obj *Series, period int) *Series
- func SMA(obj *Series, period int) *Series
- func StdDev(obj *Series, period int) *Series
- func StdDevBy(obj *Series, period int, ddof int) *Series
- func Stiffness(obj *Series, maLen, stiffLen, stiffMa int) *Series
- func Stoch(high, low, close *Series, period int) *Series
- func StochRSI(obj *Series, rsiLen int, stochLen int, maK int, maD int) *Series
- func Sum(obj *Series, period int) *Series
- func TD(obj *Series) *Series
- func TR(high *Series, low *Series, close *Series) *Series
- func UpDown(obj *Series, vtype int) *Series
- func VWMA(price *Series, vol *Series, period int) *Series
- func WMA(obj *Series, period int) *Series
- func WillR(e *BarEnv, period int) *Series
- func (s *Series) Abs() *Series
- func (s *Series) Add(obj interface{}) *Series
- func (s *Series) Append(obj interface{}) *Series
- func (s *Series) Back(num int) *Series
- func (s *Series) Cached() bool
- func (s *Series) Cut(keepNum int)
- func (s *Series) Get(i int) float64
- func (s *Series) Len() int
- func (s *Series) Max(obj interface{}) *Series
- func (s *Series) Min(obj interface{}) *Series
- func (s *Series) Mul(obj interface{}) *Series
- func (s *Series) Range(start, stop int) []float64
- func (s *Series) Set(obj interface{}) *Series
- func (s *Series) Sub(obj interface{}) *Series
- func (s *Series) To(k string, v int) *Series
- type XState
Constants ¶
const ( MinPenLen = 5 CLInit = 0 // 尚未确认有效 CLValid = 1 // 确认有效，尚未完成 CLDone = 2 // 确认有效，已完成，不会再改 ParseLevelPen = 1 ParseLevelSeg = 2 ParseLevelCentre = 3 ParseLevelTrend = 4 DefMaxPen = 1000 // 默认保存的最大笔数量 )
const ( EvtNew = 1 EvtChange = 2 EvtRemove = 3 )
Variables ¶
var ( ErrInvalidSeriesVal = errors.New("invalid val for Series") ErrGetDataOfMerged = errors.New("try get Data of merged series var") )
Functions ¶
func Cross ¶
Cross 计算最近一次交叉的距离。如果两个都变化，则两个都必须是序列。或者一个是常数一个是序列 返回值：正数上穿，负数下穿，0表示未知或重合；abs(ret) - 1表示交叉点与当前bar的距离
func RemoveFromArr ¶ added in v0.1.3
func RemoveFromArr[T comparable](arr []T, it T, num int) []T
Types ¶
type BarEnv ¶
type BarEnv struct {
TimeStart int64
TimeStop int64
Exchange string
MarketType string
Symbol string
TimeFrame string
TFMSecs int64 //周期的毫秒间隔
BarNum int
MaxCache int
VNum int
Open *Series
High *Series
Low *Series
Close *Series
Volume *Series
Info *Series
Data map[string]interface{}
}
func (*BarEnv) TrimOverflow ¶
func (e *BarEnv) TrimOverflow()
type CCentre ¶ added in v0.1.3
type CCentre struct {
*CTwoPoint
Overlap [2]float64 // 中枢重叠区间
Range [2]float64 // 中枢高低区间
Dirt float64
}
中枢
type CGraph ¶ added in v0.1.3
type CGraph struct {
Pens []*CPen
Segs []*CSeg
Trends []*CTrend
Bars []*Kline
Centres []*CCentre
OnPoint func(p *CPoint, evt int)
OnPen func(p *CPen, evt int)
OnSeg func(p *CSeg, evt int)
OnCentre func(c *CCentre, evt int)
BarNum int // 最后一个bar的序号，从1开始
ParseLevel int // 期望解析到的等级，0不限制
ParseTo int // 解析到K线的位置，序号，非索引
MaxPen int // 保存最大笔数量，默认1000
// contains filtered or unexported fields
}
type CPoint ¶ added in v0.1.3
type CPoint struct {
Graph *CGraph
Dirt float64 // 1顶分型，-1底分型
BarId int // 此bar的序号
Price float64 // 价格
StartPen *CPen
EndPen *CPen
StartSeg *CSeg
EndSeg *CSeg
StartTrend *CTrend
EndTrend *CTrend
Next *CPoint
}
type CSeg ¶ added in v0.1.3
type CSeg struct {
*CTwoPoint
Dirt float64 // 1向上 0不确定 -1向下
Feas [][2]float64 // 特征序列
InForce bool // 标记此线段是否必须以顶底分型确认
Temp bool // 标记此线段是假设的，需要再次检查是否有效
Prev *CSeg
Next *CSeg
Centre *CCentre // 此线段所属中枢，只对中间部分线段设置；一个线段只可属于一个中枢
}
线段
func (*CSeg) AddFeature ¶ added in v0.1.3
AddFeature 将笔作为特征序列
func (*CSeg) CalcFeatures ¶ added in v0.1.3
func (s *CSeg) CalcFeatures()
CalcFeatures 重新计算线段的所有特征序列，当线段结束点修改时调用此方法
func (*CSeg) SetLastFea ¶ added in v0.1.3
type Series ¶
type Series struct {
ID int
Env *BarEnv
Data []float64
Cols []*Series
Time int64
More interface{}
Subs map[string]map[int]*Series // 由此序列派生的；function：hash：object
XLogs map[int]*CrossLog // 此序列交叉记录
// contains filtered or unexported fields
}
func ADXBy ¶ added in v0.1.3
ADXBy Average Directional Index
method=0 classic ADX method=1 TradingView "ADX and DI for v4"
suggest period: 14
return [maDX, plusDI, minusDI]
func ALMA ¶ added in v0.1.6
ALMA Arnaud Legoux Moving Average
period: window size. Default: 10
sigma: Smoothing value. Default 6.0
distOff: min 0 (smoother), max 1 (more responsive). Default: 0.85
func Aroon ¶ added in v0.1.3
Aroon 阿隆指标
Reflects the distance between the highest price and the lowest price within a certain period of time. 反映了一段时间内出现最高价和最低价距离当前时间的远近。
AroonUp: (period - HighestBar(high, period+1)) / period * 100
AroonDn: (period - LowestBar(low, period+1)) / period * 100
Osc: AroonUp - AroonDn
return [AroonUp, Osc, AroonDn]
func BBANDS ¶
BBANDS Bollinger Bands 布林带指标
period: 20, stdUp: 2, stdDn: 2
return [upper, mid, lower]
func CCI ¶ added in v0.1.3
CCI Commodity Channel Index
https://www.tradingview.com/support/solutions/43000502001-commodity-channel-index-cci/
suggest period: 20
func CHOP ¶ added in v0.1.6
CHOP Choppiness Index
suggest period: 14
higher values equal more choppiness, while lower values indicate directional trending. 值越高，波动性越大，而值越低，则表示有方向性趋势。
func CMF ¶ added in v0.1.3
CMF Chaikin Money Flow
https://www.tradingview.com/scripts/chaikinmoneyflow/?solution=43000501974
suggest period: 20
func CMO ¶ added in v0.1.6
CMO Chande Momentum Oscillator
suggest period: 9
Same implementation as ta-lib For TradingView, use: CMOBy(obj, period, 1)
func CMOBy ¶ added in v0.1.6
CMOBy Chande Momentum Oscillator
suggest period: 9
maType: 0: ta-lib 1: tradingView
func CRSI ¶ added in v0.1.8
CRSI Connors RSI
suggest period:3, upDn:2, roc:100
Basically the same as TradingView
func CRSIBy ¶ added in v0.1.8
CRSIBy Connors RSI
suggest period:3, upDn:2, roc:100
vtype: 0 Calculation in TradingView method
1 Calculation in ta-lib community method:
chg = close_col / close_col.shift(1) updown = np.where(chg.gt(1), 1.0, np.where(chg.lt(1), -1.0, 0.0)) rsi = ta.RSI(close_arr, timeperiod=3) ud = ta.RSI(updown, timeperiod=2) roc = ta.ROC(close_arr, 100) crsi = (rsi + ud + roc) / 3
func CTI ¶ added in v0.1.4
CTI Correlation Trend Indicator
The Correlation Trend Indicator is an oscillator created by John Ehler in 2020. It assigns a value depending on how close prices in that range are to following a positively- or negatively-sloping straight line. Values range from -1 to 1. This is a wrapper for LinRegAdv.
suggest period: 20
func ChaikinOsc ¶ added in v0.1.3
ChaikinOsc Chaikin Oscillator
https://www.tradingview.com/support/solutions/43000501979-chaikin-oscillator/
short: 3, long: 10
func HighestBar ¶ added in v0.1.3
func KAMA ¶ added in v0.1.3
KAMA Kaufman Adaptive Moving Average
period: 10 fixed: (fast: 2, slow: 30)
func KDJBy ¶
func KDJBy(high *Series, low *Series, close *Series, period int, sm1 int, sm2 int, maBy string) *Series
KDJBy alias: stoch indicator;
period: 9, sm1: 3, sm2: 3
return (K, D, RSV)
func LinReg ¶ added in v0.1.4
LinReg Linear Regression Moving Average
Linear Regression Moving Average (LINREG). This is a simplified version of a Standard Linear Regression. LINREG is a rolling regression of one variable. A Standard Linear Regression is between two or more variables.
func MACD ¶
MACD
Internationally, init_type=0 is used, while MyTT and China mainly use init_type=1 国外主流使用init_type=0，MyTT和国内主要使用init_type=1
fast: 12, slow: 26, smooth: 9
return [macd, signal]
func PercentRank ¶ added in v0.1.8
PercentRank
calculates the percentile rank of a bar value in a data set.
func RMA ¶
RMA Relative Moving Average 相对移动均线
The difference from EMA is: both the numerator and denominator are reduced by 1 Latest value weight: 1/n
和EMA区别是：分子分母都减一 最近一个权重：1/n
func RMABy ¶
RMABy Relative Moving Average 相对移动均线
The difference from EMA is: both the numerator and denominator are reduced by 1 The most recent weight: 1/n
和EMA区别是：分子分母都减一 最近一个权重：1/n
initType: 0 initialize with SMA, 1 initialize with the first valid value
initVal defaults to Nan
initType：0使用SMA初始化，1第一个有效值初始化
initVal 默认Nan
func RMI ¶ added in v0.1.3
RMI Relative Momentum Index https://theforexgeek.com/relative-momentum-index/
period: 14, montLen: 3
func Stoch ¶ added in v0.1.3
Stoch 100 * (close - lowest(low, period)) / (highest(high, period) - lowest(low, period))
use KDJ if you want to apply SMA/RMA to this
suggest period: 14
func StochRSI ¶ added in v0.1.3
StochRSI StochasticRSI
rsiLen: 14, stochLen: 14, maK: 3, maD: 3
return [fastK, fastD]
func UpDown ¶ added in v0.1.8
UpDown
vtype: 0 TradingView (Count consecutive times)
1 classic (abs count up to 1)
func VWMA ¶ added in v0.1.3
VWMA Volume Weighted Moving Average 成交量加权平均价格
sum(price*volume)/sum(volume)
suggest period: 20
func WMA ¶ added in v0.1.4
WMA Weighted Moving Average.
the weighting factors decrease in arithmetic progression.
suggest period: 9
Source Files