Repository files navigation

This repository provides technical tools to analyze OHLCV data, along with severalTA chart functionalities. These functions are optimized for speed and utilize numpyvectorization over built-in pandas methods when possible.

• Bollinger(df=None, filename=None, interval=None, n=20, ndev=2): Bollinger Bands
• Ichimoku(df=None, filename=None, interval=None): Ichimoku Cloud
• Renko(df=None, filename=None, interval=None): Renko Chart

• atr(high, low, close, n=2): average true range from candlestick data
• cmf(df, n=2): Chaikin Money Flow of an OHLCV dataset
• double_smooth(src, n_slow, n_fast): The smoothed value of two EMAs
• ema(src, n=2): exponential moving average for a list ofsrc acrossn periods
• macd(src, slow=25, fast=13): moving average convergence/divergence ofsrc
• mmo(src, n=2): Murrey Math oscillator ofsrc
• roc(src, n=2): rate of change ofsrc acrossn periods
• rolling(src, n=2, fn=None, axis=1): rollingsum,max,min,mean, ormedian ofsrc acrossn periods
• rsi(src, n=2): relative strength index ofsrc acrossn periods
• sdev(src, n=2): standard deviation across n periods
• sma(src, n=2): simple moving average ofsrc acrossn periods
• td_sequential(src, n=2): TD sequential ofsrc acrossn periods
• tsi(src, slow=25, fast=13): true strength indicator

• area_between(line1, line2): find the area between line1 and line2
• crossover(x1, x2): find all instances of intersections between two lines
• draw_candlesticks(ax, df): add candlestick visuals to a matplotlib chart
• fill_values(averages, interval, target_len): Fill missing values with evenlyspaced samples.
  • Example: You're using 15-min candlestick data to find the 1-hour moving averageand want a value at every 15-min mark, and not every 1-hour mark.
• group_candles(df, interval=4): combine candles so instead of needing a differentdataset for each time interval, you can form time intervals using more precisedata.
  • Example: you have 15-min candlestick data but want to test a strategy basedon 1-hour candlestick data (interval=4).
• intersection(a0, a1, b0, b1): find the intersection coordinates between vectorA and vector B

# NOTE: we are using 1-hour BTC OHLCV data from 2019.01.01 00:00:00 to 2019.12.31 23:00:00fromTAcharts.utils.ohlcvimportOHLCVdf=OHLCV().btcdf.head()

fromTAcharts.indicators.bollingerimportBollingerb=Bollinger(df)b.build(n=20,ndev=2)b.plot()

fromTAcharts.indicators.ichimokuimportIchimokui=Ichimoku(df)i.build(20,60,120,30)i.plot()

fromTAcharts.indicators.renkoimportRenkor=Renko(df)r.set_brick_size(auto=True,atr_interval=2)r.build()r.plot()

• @args_to_dtype(dtype): Convert all function arguments to a specific data type

  fromTAcharts.wrappersimportargs_to_dtype# Example: `src` is converted to a list@args_to_dtype(list)defrsi(src,n=2):pass

• @pd_series_to_np_array: Convert function arguments frompd.Series tonp.array usingpd.Series.values. This wrapper is 10x quicker than using@args_to_dtype(np.array) when working with Pandas series.

  fromTAcharts.wrappersimportpd_series_to_np_array# Example: `high`, `low`, and `close` are all converted into `np.array` data types@pd_series_to_np_arraydefatr(high,low,close,n=14):pass