gavindiaz/QuanTAlib
1
0
Fork 0
mirror of https://github.com/mihakralj/QuanTAlib.git synced 2026-08-05 20:47:43 +00:00
Code Issues Packages Projects Releases Wiki Activity

SUM, MIDPOINT, MIDPRICE

Browse Source
This commit is contained in:
Miha Kralj
2022-11-14 09:52:40 -08:00
parent ddfcb6bd99
commit ee11595ffa
7 changed files with 372 additions and 166 deletions
Download Patch File Download Diff File Expand all files Collapse all files
Source/Basics/MIDPOINT_Series.cs
+44
View File
@@ -0,0 +1,44 @@
namespace QuanTAlib;
using System;
/* <summary>
MIDPOINT: Midpoint value (max+min)/2 in the given period in the series.
If period = 0 => period = full length of the series
Sources:
https://thefaqblog.com/what-is-the-midpoint-in-statistics/
</summary> */
public class MIDPOINT_Series : Single_TSeries_Indicator
{
public MIDPOINT_Series(TSeries source, int period, bool useNaN = false) : base(source, period, useNaN)
{
if (base._data.Count > 0)
{ base.Add(base._data); }
}
private readonly System.Collections.Generic.List<double> _buffer = new();
public override void Add((DateTime t, double v) TValue, bool update)
{
if (update)
{ this._buffer[this._buffer.Count - 1] = TValue.v; }
else
{ this._buffer.Add(TValue.v); }
if (this._buffer.Count > this._p && this._p != 0)
{ this._buffer.RemoveAt(0); }
double _max = TValue.v;
double _min = TValue.v;
for (int i = 0; i < this._buffer.Count; i++)
{
_max = Math.Max(this._buffer[i], _max);
_min = Math.Min(this._buffer[i], _min);
}
double _mid = (_max + _min) * 0.5;
var result = (TValue.t, this.Count < this._p - 1 && this._NaN ? double.NaN : _mid);
base.Add(result, update);
}
}
Source/Basics/MIDPRICE_Series.cs
+50
View File
@@ -0,0 +1,50 @@
namespace QuanTAlib;
using System;
/* <summary>
MIDPRICE: Midpoint price (highhest high + lowest low)/2 in the given period in the series.
If period = 0 => period = full length of the series
</summary> */
public class MIDPRICE_Series : Single_TBars_Indicator
{
public MIDPRICE_Series(TBars source, int period, bool useNaN = false) : base(source, period, useNaN)
{
if (base._bars.Count > 0)
{ base.Add(base._bars); }
}
private readonly System.Collections.Generic.List<double> _bufferhi = new();
private readonly System.Collections.Generic.List<double> _bufferlo = new();
public override void Add((DateTime t, double o, double h, double l, double c, double v) TBar, bool update)
{
if (update)
{
this._bufferhi[this._bufferhi.Count - 1] = TBar.h;
this._bufferlo[this._bufferlo.Count - 1] = TBar.l;
}
else
{
this._bufferhi.Add(TBar.h);
this._bufferlo.Add(TBar.l);
}
if (this._bufferhi.Count > this._p && this._p != 0)
{ this._bufferhi.RemoveAt(0); }
if (this._bufferlo.Count > this._p && this._p != 0)
{ this._bufferlo.RemoveAt(0); }
double _max = TBar.h;
double _min = TBar.l;
for (int i = 0; i < this._bufferhi.Count; i++)
{
_max = Math.Max(this._bufferhi[i], _max);
_min = Math.Min(this._bufferlo[i], _min);
}
double _mid = (_max + _min) * 0.5;
var result = (TBar.t, this.Count < this._p - 1 && this._NaN ? double.NaN : _mid);
base.Add(result, update);
}
}
Source/Basics/SUM_Series.cs
+35
View File
@@ -0,0 +1,35 @@
namespace QuanTAlib;
using System;
/* <summary>
SUM: Cumulative Sum (aka Running Total)
SUM across a period provides a rolling sum of all values across the period.
If SUM values would be divided with period, the output would be SMA()
Sources:
https://en.wikipedia.org/wiki/CUSUM
</summary> */
public class SUM_Series : Single_TSeries_Indicator
{
public SUM_Series(TSeries source, int period, bool useNaN = false) : base(source, period, useNaN)
{
if (base._data.Count > 0) { base.Add(base._data); }
}
private readonly System.Collections.Generic.List<double> _buffer = new();
public override void Add((System.DateTime t, double v) TValue, bool update)
{
if (update) { _buffer[_buffer.Count - 1] = TValue.v; }
else { _buffer.Add(TValue.v); }
if (_buffer.Count > this._p && this._p != 0) { _buffer.RemoveAt(0); }
double _sum = 0;
for (int i = 0; i < _buffer.Count; i++) { _sum += _buffer[i]; }
var result = (TValue.t, (this.Count < this._p - 1 && this._NaN) ? double.NaN : _sum);
base.Add(result, update);
}
}
Source/Trends/ALMA_Series.cs
+47 -45
View File
@@ -14,52 +14,54 @@ Sources:
https://phemex.com/academy/what-is-arnaud-legoux-moving-averages
https://www.prorealcode.com/prorealtime-indicators/alma-arnaud-legoux-moving-average/
TODO: Discrepancy with Pandas-TA (but passes the validation with Skender.GetAlma)
</summary> */
public class ALMA_Series : Single_TSeries_Indicator
{
private readonly System.Collections.Generic.List<double> _buffer = new();
private readonly double[] _weight;
private double _norm;
private readonly double _offset, _sigma;
public ALMA_Series(TSeries source, int period, double offset = 0.85, double sigma = 6.0, bool useNaN = false)
: base(source, period, useNaN)
{
_offset = offset;
_sigma = sigma;
_weight = new double[period];
if (this._data.Count > 0) { base.Add(this._data); }
}
public override void Add((System.DateTime t, double v) TValue, bool update)
{
if (update) { this._buffer[this._buffer.Count - 1] = TValue.v; }
else { this._buffer.Add(TValue.v); }
if (this._buffer.Count > this._p) { this._buffer.RemoveAt(0); }
if (this._buffer.Count <= _p) { calc_weights(); }
double _weightedSum = 0;
for (int i = 0; i < this._buffer.Count; i++) { _weightedSum += _weight[i] * _buffer[i]; }
double _alma = _weightedSum / _norm;
var ret = (TValue.t, this.Count < this._p - 1 && this._NaN ? double.NaN : _alma);
base.Add(ret, update);
}
private void calc_weights()
{
int _len = this._buffer.Count;
_norm = 0;
double _m = _offset * (_len - 1);
double _s = _len / _sigma;
for (int i = 0; i < _len; i++)
{
double _wt = Math.Exp(-((i - _m) * (i - _m)) / (2 * _s * _s));
_weight[i] = _wt;
_norm += _wt;
}
}
{
private readonly System.Collections.Generic.List<double> _buffer = new();
private readonly double[] _weight;
private double _norm;
private readonly double _offset, _sigma;
public ALMA_Series(TSeries source, int period, double offset = 0.85, double sigma = 6.0, bool useNaN = false)
: base(source, period, useNaN)
{
_offset = offset;
_sigma = sigma;
_weight = new double[period];
if (this._data.Count > 0) { base.Add(this._data); }
}
public override void Add((System.DateTime t, double v) TValue, bool update)
{
if (update) { this._buffer[this._buffer.Count - 1] = TValue.v; }
else { this._buffer.Add(TValue.v); }
if (this._buffer.Count > this._p) { this._buffer.RemoveAt(0); }
if (this._buffer.Count <= _p) { calc_weights(); }
double _weightedSum = 0;
for (int i = 0; i < this._buffer.Count; i++) { _weightedSum += _weight[i] * _buffer[i]; }
double _alma = _weightedSum / _norm;
var ret = (TValue.t, this.Count < this._p - 1 && this._NaN ? double.NaN : _alma);
base.Add(ret, update);
}
private void calc_weights()
{
int _len = this._buffer.Count;
_norm = 0;
double _m = _offset * (_len - 1);
double _s = _len / _sigma;
for (int i = 0; i < _len; i++)
{
double _wt = Math.Exp(-((i - _m) * (i - _m)) / (2 * _s * _s));
_weight[i] = _wt;
_norm += _wt;
}
}
}
Tests/Validations/Pandas_TA.cs
+166 -118
View File
@@ -6,123 +6,171 @@ using Python.Included;
namespace Validations;
public class PandasTA : IDisposable
{
private GBM_Feed bars;
private Random rnd = new();
private int period;
private string OStype;
private dynamic np;
private dynamic ta;
private dynamic df;
public PandasTA()
{
bars = new(5000);
period = rnd.Next(28) + 3;
// Checking the host OS and setting PythonDLL accordingly
OStype = Environment.OSVersion.ToString();
if (OStype == "Unix 13.1.0")
OStype = @"/usr/local/Cellar/python@3.10/3.10.8/Frameworks/Python.framework/Versions/3.10/lib/libpython3.10.dylib";
else OStype = Path.GetFullPath(".") + @"\python-3.10.0-embed-amd64\python310.dll";
Installer.InstallPath = Path.GetFullPath(".");
Installer.SetupPython().Wait();
Installer.TryInstallPip();
Installer.PipInstallModule("pandas-ta");
Runtime.PythonDLL = OStype;
PythonEngine.Initialize();
np = Py.Import("numpy");
ta = Py.Import("pandas_ta");
string[] cols = { "open", "high", "low", "close", "volume" };
double[,] ary = new double[bars.Count, 5];
for (int i = 0; i < bars.Count; i++)
{
ary[i, 0] = bars.Open[i].v;
ary[i, 1] = bars.High[i].v;
ary[i, 2] = bars.Low[i].v;
ary[i, 3] = bars.Close[i].v;
ary[i, 4] = bars.Volume[i].v;
}
df = ta.DataFrame(data: np.array(ary), index: np.array(bars.Close.t), columns: np.array(cols));
}
public void Dispose()
{
PythonEngine.Shutdown();
}
[Fact]
void SMA()
{
SMA_Series QL = new(bars.Close, period, false);
var pta = df.ta.sma(close: df.close, length: period);
Assert.Equal(Math.Round((double)pta.tail(1), 7), Math.Round(QL.Last().v, 7));
}
[Fact]
void EMA()
{
EMA_Series QL = new(bars.Close, period, false);
var pta = df.ta.ema(close: df.close, length: period);
Assert.Equal(Math.Round((double)pta.tail(1), 7), Math.Round(QL.Last().v, 7));
}
[Fact]
void TEMA()
{
TEMA_Series QL = new(bars.Close, period, false);
var pta = df.ta.tema(close: df.close, length: period);
Assert.Equal(Math.Round((double)pta.tail(1), 7), Math.Round(QL.Last().v, 7));
}
[Fact]
void ENTP()
{
ENTP_Series QL = new(bars.Close, period, useNaN: false);
var pta = df.ta.entropy(close: df.close, length: period);
Assert.Equal(Math.Round((double)pta.tail(1), 7), Math.Round(QL.Last().v, 7));
}
[Fact]
void WMA()
{
WMA_Series QL = new(bars.Close, period, false);
var pta = df.ta.wma(close: df.close, length: period);
Assert.Equal(Math.Round((double)pta.tail(1), 7), Math.Round(QL.Last().v, 7));
}
[Fact]
void DEMA()
{
DEMA_Series QL = new(bars.Close, period, false);
var pta = df.ta.dema(close: df.close, length: period);
Assert.Equal(Math.Round((double)pta.tail(1), 7), Math.Round(QL.Last().v, 7));
}
[Fact]
void BIAS()
{
BIAS_Series QL = new(bars.Close, period, false);
var pta = df.ta.bias(close: df.close, length: period);
Assert.Equal(Math.Round((double)pta.tail(1), 7), Math.Round(QL.Last().v, 7));
}
[Fact]
void KURT()
{
KURT_Series QL = new(bars.Close, period, useNaN: false);
var pta = df.ta.kurtosis(close: df.close, length: period);
Assert.Equal(Math.Round((double)pta.tail(1), 4), Math.Round(QL.Last().v, 4));
}
[Fact]
void MAD()
{
MAD_Series QL = new(bars.Close, period, useNaN: false);
var pta = df.ta.mad(close: df.close, length: period);
Assert.Equal(Math.Round((double)pta.tail(1), 7), Math.Round(QL.Last().v, 7));
{
private GBM_Feed bars;
private Random rnd = new();
private int period;
private string OStype;
private dynamic np;
private dynamic ta;
private dynamic df;
public PandasTA()
{
bars = new(5000);
period = rnd.Next(28) + 3;
// Checking the host OS and setting PythonDLL accordingly
OStype = Environment.OSVersion.ToString();
if (OStype == "Unix 13.1.0")
OStype = @"/usr/local/Cellar/python@3.10/3.10.8/Frameworks/Python.framework/Versions/3.10/lib/libpython3.10.dylib";
else OStype = Path.GetFullPath(".") + @"\python-3.10.0-embed-amd64\python310.dll";
Installer.InstallPath = Path.GetFullPath(".");
Installer.SetupPython().Wait();
Installer.TryInstallPip();
//Installer.PipInstallModule("pandas-ta");
Installer.PipInstallModule("git+https://github.com/twopirllc/pandas-ta@development");
Runtime.PythonDLL = OStype;
PythonEngine.Initialize();
np = Py.Import("numpy");
ta = Py.Import("pandas_ta");
string[] cols = { "open", "high", "low", "close", "volume" };
double[,] ary = new double[bars.Count, 5];
for (int i = 0; i < bars.Count; i++)
{
ary[i, 0] = bars.Open[i].v;
ary[i, 1] = bars.High[i].v;
ary[i, 2] = bars.Low[i].v;
ary[i, 3] = bars.Close[i].v;
ary[i, 4] = bars.Volume[i].v;
}
df = ta.DataFrame(data: np.array(ary), index: np.array(bars.Close.t), columns: np.array(cols));
}
public void Dispose()
{
PythonEngine.Shutdown();
}
[Fact]
void HL2()
{
var pta = df.ta.hl2(high: df.high, low: df.low);
Assert.Equal(Math.Round((double)pta.tail(1), 7), Math.Round(bars.HL2.Last().v, 7));
}
[Fact]
void HLC3()
{
var pta = df.ta.hlc3(high: df.high, low: df.low, close: df.close);
Assert.Equal(Math.Round((double)pta.tail(1), 7), Math.Round(bars.HLC3.Last().v, 7));
}
[Fact]
void OHLC4()
{
var pta = df.ta.ohlc4(open: df.open, high: df.high, low: df.low, close: df.close);
Assert.Equal(Math.Round((double)pta.tail(1), 7), Math.Round(bars.OHLC4.Last().v, 7));
}
[Fact]
void KAMA()
{
KAMA_Series QL = new(bars.Close, period);
var pta = df.ta.kama(close: df.close, length: period);
Assert.Equal(Math.Round((double)pta.tail(1), 7), Math.Round(QL.Last().v, 7));
}
/*
[Fact]
void ALMA()
{
ALMA_Series QL = new(bars.Close, period: period, offset: 0.85, sigma: 6.0, false);
var pta = df.ta.alma(close: df.close, length: period, distribution_offset: 0.85, sigma: 6.0);
Assert.Equal(Math.Round((double)pta.tail(1), 7), Math.Round(QL.Last().v, 7));
}
*/
[Fact]
void HMA()
{
HMA_Series QL = new(bars.Close, period, false);
var pta = df.ta.hma(close: df.close, length: period);
Assert.Equal(Math.Round((double)pta.tail(1), 7), Math.Round(QL.Last().v, 7));
}
[Fact]
void SMA()
{
SMA_Series QL = new(bars.Close, period, false);
var pta = df.ta.sma(close: df.close, length: period);
Assert.Equal(Math.Round((double)pta.tail(1), 7), Math.Round(QL.Last().v, 7));
}
[Fact]
void EMA()
{
EMA_Series QL = new(bars.Close, period, false);
var pta = df.ta.ema(close: df.close, length: period);
Assert.Equal(Math.Round((double)pta.tail(1), 7), Math.Round(QL.Last().v, 7));
}
[Fact]
void TEMA()
{
TEMA_Series QL = new(bars.Close, period, false);
var pta = df.ta.tema(close: df.close, length: period);
Assert.Equal(Math.Round((double)pta.tail(1), 7), Math.Round(QL.Last().v, 7));
}
[Fact]
void ENTP()
{
ENTP_Series QL = new(bars.Close, period, useNaN: false);
var pta = df.ta.entropy(close: df.close, length: period);
Assert.Equal(Math.Round((double)pta.tail(1), 7), Math.Round(QL.Last().v, 7));
}
[Fact]
void WMA()
{
WMA_Series QL = new(bars.Close, period, false);
var pta = df.ta.wma(close: df.close, length: period);
Assert.Equal(Math.Round((double)pta.tail(1), 7), Math.Round(QL.Last().v, 7));
}
[Fact]
void DEMA()
{
DEMA_Series QL = new(bars.Close, period, false);
var pta = df.ta.dema(close: df.close, length: period);
Assert.Equal(Math.Round((double)pta.tail(1), 7), Math.Round(QL.Last().v, 7));
}
[Fact]
void BIAS()
{
BIAS_Series QL = new(bars.Close, period, false);
var pta = df.ta.bias(close: df.close, length: period);
Assert.Equal(Math.Round((double)pta.tail(1), 7), Math.Round(QL.Last().v, 7));
}
[Fact]
void KURT()
{
KURT_Series QL = new(bars.Close, period, useNaN: false);
var pta = df.ta.kurtosis(close: df.close, length: period);
Assert.Equal(Math.Round((double)pta.tail(1), 4), Math.Round(QL.Last().v, 4));
}
[Fact]
void MAD()
{
MAD_Series QL = new(bars.Close, period, useNaN: false);
var pta = df.ta.mad(close: df.close, length: period);
Assert.Equal(Math.Round((double)pta.tail(1), 7), Math.Round(QL.Last().v, 7));
}
}
Tests/Validations/TA_LIB.cs
+27
View File
@@ -85,6 +85,33 @@ public class TA_LIB
}
[Fact]
public void SUM()
{
SUM_Series QL = new(bars.Close, period, false);
Core.Sum(inclose, 0, bars.Count - 1, TALIB, out int outBegIdx, out _, period);
Assert.Equal(Math.Round(TALIB[TALIB.Length - outBegIdx - 1], 6, MidpointRounding.AwayFromZero), Math.Round(QL.Last().v, 6, MidpointRounding.AwayFromZero));
}
[Fact]
public void MIDPRICE()
{
MIDPRICE_Series QL = new(bars, period, false);
Core.MidPrice(inhigh, inlow, 0, bars.Count - 1, TALIB, out int outBegIdx, out _, period);
Assert.Equal(Math.Round(TALIB[TALIB.Length - outBegIdx - 1], 6, MidpointRounding.AwayFromZero), Math.Round(QL.Last().v, 6, MidpointRounding.AwayFromZero));
}
[Fact]
public void MIDPOINT()
{
MIDPOINT_Series QL = new(bars.Close, period, false);
Core.MidPoint(inclose, 0, bars.Count - 1, TALIB, out int outBegIdx, out _, period);
Assert.Equal(Math.Round(TALIB[TALIB.Length - outBegIdx - 1], 6, MidpointRounding.AwayFromZero), Math.Round(QL.Last().v, 6, MidpointRounding.AwayFromZero));
}
[Fact]
public void TRIMA()
{
TRIMA_Series QL = new(bars.Close, period, false);
docs/readme.md
+3 -3
View File
@@ -43,11 +43,11 @@ See [Getting Started](https://github.com/mihakralj/QuanTAlib/blob/main/Docs/gett
| ✔️ OHL3 - (Open+High+Low)/3 | `.OHL3` |||
| ⭐ OHLC4 - Average Price | `.OHLC4` | AVGPRICE |️ GetBaseQuote |
| ⭐ HLCC4 - Weighted Price | `.HLCC4` | WCLPRICE ||
| ⭐ MIDPOINT - Midpoint value | `MIDPOINT_Series` | MIDPOINT ||
| ⭐ MIDPRICE - Midpoint price | `MIDPRICE_Series` | MIDPRICE ||
| ⭐ MAX - Max value | `MAX_Series` | MAX ||
| ⭐ MIN - Min value | `MIN_Series` | MIN ||
| ⛔ MID - Midpoint value || MIDPOINT ||
| ⛔ MIDP - Midpoint price || MIDPRICE ||
| ⛔ SUM - Summation || SUM ||
| ⭐ SUM - Summation | `SUM_Series` | SUM ||
| ⭐ ADD - Addition | `ADD_Series` | ADD ||
| ⭐ SUB - Subtraction | `SUB_Series` | SUB ||
| ⭐ MUL - Multiplication | `MUL_Series` | MUL ||