Technical Analysis (ta.py)
ta.py is a Python package for dealing with financial technical analysis.
Installation
pip
Use the package manager pip to install ta.py.
pip install ta_py
Usage
import ta_py as ta;
Examples
Moving Averages
- Simple Moving Average
- Smoothed Moving Average
- Weighted Moving Average
- Exponential Moving Average
- Hull Moving Average
- Least Squares Moving Average
- Volume Weighted Moving Average
- Wilder's Smoothing Moving Average
- Parabolic Weighted Moving Average
- Hyperbolic Weighted Moving Average
- Kaufman Adaptive Moving Average
Indicators
- Moving Average Convergence / Divergence
- Relative Strength Index
- Wilder's Relative Strength Index
- True Strength Index
- Balance Of Power
- Force Index
- Accumulative Swing Index
- Alligator Indicator
- Williams %R
- Stochastics
- Fibonacci Retracement
- Bollinger Bandwidth
- Ichimoku Cloud
- Average True Range
- Aroon Up
- Aroon Down
- Money Flow Index
- Rate Of Change
- Coppock Curve
- Know Sure Thing
- On-Balance Volume
- Volume-Weighted Average Price
- Fractals
- Momentum
Oscillators
- Alligator Oscillator
- Chande Momentum Oscillator
- Chaikin Oscillator
- Aroon Oscillator
- Awesome Oscillator
- Accelerator Oscillator
Bands
Statistics
- Standard Deviation
- Variance
- Inverse Normal Distribution
- Monte Carlo Simulation
- Percentile
- Correlation
- Percentage Difference
- Expected Return
- Drawdown
- Median
- Recent High
- Recent Low
- Median Absolute Deviation
- Average Absolute Deviation
- Sum Squared Differences
- Normalize
- Denormalize
- Standardize
- K-means Clustering
Chart Types
Experimental
Moving Averages
Simple Moving Average (SMA)
data = [1, 2, 3, 4, 5, 6, 10];
length = 6; # default = 14
ta.sma(data, length);
# output (array)
# [3.5, 5]
Smoothed Moving Average (SMMA)
data = [1, 2, 3, 4, 5, 6, 10];
length = 5; # default = 14
ta.smma(data, length);
# output (array)
# [3.4, 4.92]
Weighted Moving Average (WMA)
data = [69, 68, 66, 70, 68];
length = 4; # default = 14
ta.wma(data, length);
# output (array)
# [68.3, 68.2]
Exponential Moving Average (EMA)
data = [1, 2, 3, 4, 5, 6, 10];
length = 6; # default = 12
ta.ema(data, length);
# output (array)
# [3.5, 5.357]
Hull Moving Average
data = [6, 7, 5, 6, 7, 4, 5, 7];
length = 6; # default = 14
ta.hull(data, length);
# output (array)
# [4.76, 5.48]
Least Squares Moving Average (LSMA)
data = [5, 6, 6, 3, 4, 6, 7];
length = 6; # default = 25
ta.lsma(data, length);
# output (array)
# [4.714, 5.761]
Volume Weighted Moving Average (VWMA)
data = [[1, 59], [1.1, 82], [1.21, 27], [1.42, 73], [1.32, 42]]; # [price, volume (quantity)]
length = 4; # default = 20
ta.vwma(data, length);
# output (array)
# [1.185, 1.259]
Wilder's Smoothing Moving Average
data = [1, 2, 3, 4, 5, 6, 10];
length = 6; # default = 14
ta.wsma(data, length);
# output (array)
# [3.5, 4.58]
Parabolic Weighted Moving Average
data = [17, 26, 23, 29, 20];
length = 4; # default = 14
ta.pwma(data, length);
# output (array)
# [24.09, 25.18]
Hyperbolic Weighted Moving Average
data = [54, 51, 86, 42, 47];
length = 4; # default = 14
ta.hwma(data, length);
# output (array)
# [56.2, 55.0]
Kaufman Adaptive Moving Average (KAMA)
data = [8, 7, 8, 9, 7, 9];
length1 = 2; # default = 10
length2 = 4; # default = 2
length3 = 8; # default = 30
ta.kama(data, length1, length2, length3);
# output (array)
# [8, 8.64, 8.57, 8.57]
Indicators
Moving Average Convergence / Divergence (MACD)
data = [1, 2, 3, 4, 5, 6, 14];
length1 = 3; # default = 12
length2 = 6; # default = 26
ta.macd(data, length1, length2);
# output (array)
# [1.5, 3]
Relative Strength Index (RSI)
data = [1, 2, 3, 4, 5, 6, 7, 5];
length = 6; # default = 14
ta.rsi(data, length);
# output (array)
# [100, 100, 66.667]
Wilder's Relative Strength Index
data = [1, 2, 3, 4, 5, 6, 7, 5, 6];
length = 6; # default = 14
ta.wrsi(data, length);
# output (array)
# [100, 71.43, 75.61]
True Strength Index (TSI)
data = [1.32, 1.27, 1.42, 1.47, 1.42, 1.45, 1.59];
longlength = 3; # default = 25
shortlength = 2; # default = 13
signallength = 2; # default = 13
ta.tsi(data, longlength, shortlength, signallength);
# output (array)
# [[0.327, 0.320], [0.579, 0.706]]
# [strength line, signal line]
Balance Of Power
data = [[4, 5, 4, 5], [5, 6, 5, 6], [6, 8, 5, 6]]; # [open, high, low, close]
length = 2; # default = 14
ta.bop(data, length);
# output (array)
# [1, 0.5]
Force Index
data = [[1.4, 200], [1.5, 240], [1.1, 300], [1.2, 240], [1.5, 400]]; # [close, volume]
length = 4; # default = 13
ta.fi(data, length);
# output (array)
# [0.0075]
Accumulative Swing Index
data = [[7, 6, 4], [9, 7, 5], [9, 8, 6]]; # [high, close, low]
ta.asi(data);
# output (array)
# [0, -12.5]
Alligator Indicator
data = [8,7,8,9,7,8,9,6,7,8,6,8,10,8,7,9,8,7,9,6,7,9];
# defaults shown
jawlength = 13;
teethlength = 8;
liplength = 5;
jawshift = 8;
teethshift = 5;
lipshift = 3;
ta.alligator(data, jawlength, teethlength, liplength, jawshift, teethshift, lipshift);
# output (array)
# [jaw, teeth, lips]
Williams %R
data = [2, 1, 3, 1, 2];
length = 3; # default = 14
ta.pr(data, length);
# output (array)
# [-0, -100, -50]
Stochastics
data = [[3,2,1], [2,2,1], [4,3,1], [2,2,1]]; # [high, close, low]
length = 2; # default = 14
smoothd = 1; # default = 3
smoothk = 1; # default = 3
ta.stoch(data, length, smoothd, smoothk);
# output (array)
# [[66.667, 66.667], [33.336, 33.336]]
# [kline, dline]
Fibonacci Retracement
start = 1;
end = 2;
ta.fib(start, end);
# output (array)
# [1, 1.236, 1.382, 1.5, 1.618, 1.786, 2, 2.618, 3.618, 4.618, 5.236]
Bollinger Bandwidth
data = [1, 2, 3, 4, 5, 6];
length = 5; # default = 14
deviations = 2; # default = 1
ta.bandwidth(data, length, deviations);
# output (array)
# [1.886, 1.344]
Ichimoku Cloud
data = [[6, 3, 2], [5, 4, 2], [5, 4, 3], [6, 4, 3], [7, 6, 4], [6, 5, 3]]; # [high, close, low]
length1 = 9; # default = 9
length2 = 26; # default = 26
length3 = 52; # default = 52
displacement = 26; # default = 26
ta.ichimoku(data, length1, length2, length3, displacement);
# output (array)
# [conversion line, base line, leading span A, leading span B, lagging span]
Average True Range (ATR)
data = [[3,2,1], [2,2,1], [4,3,1], [2,2,1]]; # [high, close, low]
length = 3; # default = 14
ta.atr(data, length);
# output (array)
# [2, 1.667, 2.111, 1.741]
Aroon Up
data = [5, 4, 5, 2];
length = 3; # default = 10
ta.aroon_up(data, length);
# output (array)
# [100.0, 50.0]
Aroon Down
data = [2, 5, 4, 5];
length = 3; # default = 10
ta.aroon_down(data, length);
# output (array)
# [0.0, 50.0]
Money Flow Index
data = [[19, 13], [14, 38], [21, 25], [32, 17]]; # [buy volume, sell volume]
length = 3; # default = 14
ta.mfi(data, length);
# output (array)
# [41.54, 45.58]
Rate Of Change
data = [1, 2, 3, 4];
length = 3; # default = 14
ta.roc(data, length);
# output (array)
# [2, 1]
Coppock Curve
data = [3, 4, 5, 3, 4, 5, 6, 4, 7, 5, 4, 7, 5];
length1 = 4; # (ROC period 1) default = 11
length2 = 6; # (ROC period 2) default = 14
length3 = 5; # (WMA smoothing period) default = 10
ta.cop(data, length1, length2, length3);
# output (array)
# [0.376, 0.237]
Know Sure Thing
data = [8, 6, 7, 6, 8, 9, 7, 5, 6, 7, 6, 8, 6, 7, 6, 8, 9, 9, 8, 6, 4, 6, 5, 6, 7, 8, 9];
# roc sma #1
r1 = 5; # default = 10
s1 = 5; # default = 10
# roc sma #2
r2 = 7; # default = 15
s2 = 5; # default = 10
# roc sma #3
r3 = 10; # default = 20
s3 = 5; # default = 10
# roc sma #4
r4 = 15; # default = 30
s4 = 7; # default = 15
# signal line
sig = 4; # default = 9
ta.kst(data, r1, s1, r2, s2, r3, s3, r4, s4, sig);
# output (array)
# [[-0.68, -0.52], [-0.29, -0.58], [0.35, -0.36]]
# [kst line, signal line]
On-Balance Volume
data = [[25200, 10], [30000, 10.15], [25600, 10.17], [32000, 10.13]]; # [asset volume, close price]
ta.obv(data);
# output (array)
# [0, 30000, 55600, 23600]
Volume-Weighted Average Price
data = [[127.21, 89329], [127.17, 16137], [127.16, 23945]]; # [average price, volume (quantity)]
length = 2; # default = data.length
ta.vwap(data, length);
# output (array)
# [127.204, 127.164]
Fractals
data = [[7,6],[8,6],[9,6],[8,5],[7,4],[6,3],[7,4],[8,5]]; # [high, low]
ta.fractals(data);
# output (array, same length as input)
# [[false, false],[false,false],[true,false],[false,false],[false,false],[false,true],[false,false],[false,false]]
# [upper fractal, lower fractal]
Momentum
data = [1, 1.1, 1.2, 1.24, 1.34];
length = 4; # default = 10
percentage = false; # default = false (true returns percentage)
ta.mom(data, length, percentage);
# output (array)
# [0.24, 0.24]
Oscillators
Alligator Oscillator
data = [8,7,8,9,7,8,9,6,7,8,6,8,10,8,7,9,8,7,9,6,7,9];
# defaults shown
jawlength = 13;
teethlength = 8;
liplength = 5;
jawshift = 8;
teethshift = 5;
lipshift = 3;
ta.gator(data, jawlength, teethlength, liplength, jawshift, teethshift, lipshift);
# output (array)
# [upper gator, lower gator]
Chande Momentum Oscillator
data = [1, 1.2, 1.3, 1.3, 1.2, 1.4];
length = 4; # default = 9
ta.mom_osc(data, length);
# output (array)
# [0.0, 3.85]
Chaikin Oscillator
data = [[2,3,4,6],[5,5,5,4],[5,4,3,7],[4,3,3,4],[6,5,4,6],[7,4,3,6]]; # [high, close, low, volume]
length1 = 2; # default = 3
length2 = 4; # default = 10
ta.chaikin_osc(data, length1, length2);
# output (array)
# [-1.667, -0.289, -0.736]
Aroon Oscillator
data = [2, 5, 4, 5];
length = 3; # default = 25
ta.aroon_osc(data, length);
# output (array)
# [50.0, 50.0]
Awesome Oscillator
data = [[6, 5], [8, 6], [7, 4], [6, 5], [7, 6], [9, 8]]; # [high, low]
shortlength = 2; # default = 5
longlength = 5; # default = 35
ta.ao(data, shortlength, longlength);
# output (array)
# [0, 0.9]
Accelerator Oscillator
data = [[6, 5], [8, 6], [7, 4], [6, 5], [7, 6], [9, 8]]; # [high, low]
shortlength = 2; # default = 5
longlength = 4; # default = 35
ta.ac(data, shortlength, longlength);
# output (array)
# [-5.875, -6.125, -6.5]
Bands
Bollinger Bands
data = [1, 2, 3, 4, 5, 6];
length = 5; # default = 14
deviations = 2; # default = 1
ta.bands(data, length, deviations);
# output (array)
# [[5.828, 3, 0.172], [6.828, 4, 1.172]]
# [upper band, middle band, lower band]
Keltner Channels
data = [[3,2,1], [2,2,1], [4,3,1], [2,2,1], [3,3,1]]; # [high, close, low]
length = 5; # default = 14
deviations = 1; # default = 1
ta.keltner(data, length, deviations);
# output (array)
# [[3.93, 2.06, 0.20]]
# [upper band, middle band, lower band]
Donchian Channels
data = [[6, 2], [5, 2], [5, 3], [6, 3], [7, 4], [6, 3]]; # [high, low]
length = 5; # default = 20
ta.don(data, length);
# output (array)
# [[7, 4.5, 2], [7, 4.5, 2]]
# [upper band, base line, lower band]
Envelope
data = [6,7,8,7,6,7,8,7,8,7,8,7,8];
length = 11, # default = 10
percentage = 0.05; # default = 0.005
ta.envelope(data, length, percentage);
# output (array)
# [[7.541, 7.182, 6.823], [7.636, 7.273, 6.909]]
# [upper band, base line, lower band]
Statistics
Standard Deviation
data = [1, 2, 3];
length = 3; # default = data.length
ta.std(data, length);
# output (float)
# 0.81649658092773
Variance
data = [6, 7, 2, 3, 5, 8, 6, 2];
length = 7; # default = len(data)
ta.variance(data, length);
# output (array)
# [3.918, 5.061]
Inverse Normal Distribution
data = 0.4732;
ta.normsinv(data);
# output (float)
# -0.06722824471054376
Monte Carlo Simulation
data = [6, 4, 7, 8, 5, 6];
length = 2; # default = 50
simulations = 100; # default = 1000
percentile = 0.5; # default = -1 (returns all raw simulations)
ta.sim(data, length, simulations, percentile)
# output (array)
# [6, 4, 7, 8, 5, 6, 5.96, 5.7]
Percentile
data = [[6,4,7], [5,3,6], [7,5,8]];
percentile = 0.5;
ta.percentile(data, percentile);
# output (array)
# [6, 4, 7]
Correlation
data1 = [1, 2, 3, 4, 5, 2];
data2 = [1, 3, 2, 4, 6, 3];
ta.cor(data1, data2);
# output (float)
# 0.8808929232684737
Percentage Difference
newval = 0.75;
oldval = 0.5;
ta.dif(newval, oldval);
# output (float)
# 0.5
Expected Return
data = [0.02, -0.01, 0.03, 0.05, -0.03]; # historical return data
ta.er(data);
# output (float)
# 0.0119
Drawdown
data = [1, 2, 3, 4, 2, 3];
ta.drawdown([1,2,3,4,2,3]);
# output (float)
# -0.5
Median
data = [4, 6, 3, 1, 2, 5];
length = 4; # default = data.length
ta.median(data, length);
# output (array)
# [3, 2, 2]
Recent High
data = [4,5,6,7,8,9,8,7,8,9,10,3,2,1];
lookback = 3; # No higher values after 3 periods? resets after each new high
ta.recent_high(data, lookback);
# output (dictionary)
# {'index': 10, 'value': 10}
Recent Low
data = [1,4,5,6,4,3,2,3,4,3,5,7,8,8,5];
lookback = 4; # No lower values after 4 periods? resets after each new low
ta.recent_low(data, lookback);
# output (dictionary)
# {'index': 6, 'value': 2}
Median Absolute Deviation
data = [3, 7, 5, 4, 3, 8, 9];
length = 6; # default = len(data)
ta.mad(data, length);
# output (array)
# [1, 2]
Average Absolute Deviation
data = [4, 6, 8, 6, 8, 9, 10, 11];
length = 7; # default = len(data)
ta.aad(data, length);
# output (array)
# [1.673, 1.469]
Sum Squared Differences
data = [7, 6, 5, 7, 9, 8, 3, 5, 4];
length = 7; # default = data.length
ta.ssd(data, length);
# output (array)
# [4.87, 4.986, 5.372]
Normalize
data = [5,4,9,4];
margin = 0.1; # margin % (default = 0)
ta.normalize(data, margin);
# output (array)
# [0.22, 0.06, 0.86, 0.06]
Denormalize
data = [5,4,9,4]; # original data || [highest, lowest]
norm = [0.22, 0.06, 0.86, 0.06, 0.44]; # normalized data
margin = 0.1; # margin % (default = 0)
ta.denormalize(data, norm, margin);
# output (array)
# [5 ,4, 9, 4, 6.4]
Standardize
data = [6,4,6,8,6];
ta.standardize(data);
# output (array)
# [0, -1.581, 0, 1.581, 0]
K-means Clustering
data = [2, 3, 4, 5, 3, 5, 7, 8, 6, 8, 6, 4, 2, 6];
length = 4;
ta.kmeans(data, length);
# output (array)
# [[ 4, 5, 5, 4 ], [ 7, 6, 6, 6 ], [ 8, 8 ], [ 2, 3, 3, 2 ]]
Chart types
Heikin Ashi
data = [[3, 4, 2, 3], [3, 6, 3, 5], [5, 5, 2, 3]]; # [open, high, low, close]
ta.ha(data);
# output (array)
# [open, high, low, close]
# first 7-10 candles are unreliable
Renko
data = [[8, 6], [9, 7], [9, 8]]; # [high, low]
bricksize = 3;
ta.ren(data, bricksize);
# output (array)
# [open, high, low, close]
Experimental Functions
Support Line
data = [4,3,2,5,7,6,5,4,7,8,5,4,6,7,5];
start = {"index": 2, "value": 2}; # default = recent_low(data, 25)
support = ta.support(data, start);
# output (dictionary)
# ['calculate'] = function(x) // calculates line at position x from start['index'] (= 0)
# ['slope'] = delta y per x
# ['lowest'] = lowest (start) value at x = 0
# ['index'] = (start) index of lowest value
# to get the line at the current candle / chart period
current = support['calculate'](len(data)-support['index']);
Resistance Line
data = [5,7,5,5,4,6,5,4,6,5,4,3,2,4,3,2,1];
start = {"index": 1, "value": 7}; # default = recent_high(data, 25)
resistance = ta.resistance(data, start);
# output (dictionary)
# ['calculate'] = function(x) // calculates line at position x from start['index'] (= 0)
# ['slope'] = delta y per x
# ['highest'] = highest (start) value
# ['index'] = (start) index of highest value
# to get the line at the current candle / chart period
current = resistance['calculate'](len(data)-resistance['index']);
Contributing
Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.
Please make sure to update tests as appropriate.