python_storage_timeline_ta_indicators

This library reads, cleans and sorts data acquired from web database. Values in a storage expected to be in a json format with a following structure:

{"time":1684968241256,"value":"{\"r\":[49208.040463174424,730195614842.3989]}"}

Class data aggregation takes link as an argument to create instance of a database. The data from the database could be updated manualy via update function.

To calculate ta indicators function print_indicators should be used that takes next arguments:

time in "2D", "4S" or "W" format that denotes length of even interwals time series will be divided.

length takes int input and stands for number of intervals that will be used to calculate the indicator.

indicators takes one (or multiple) of 'RSI', 'STOCH', 'STOCHRSI', 'ADX', 'MACD', 'WILLIAMS', 'CCI', 'ATR', 'HIGHLOW', 'ULTOSC', 'ULTOSC', 'ROC', 'MA'. If is None calculates all of listed.

IMPORTANT Some indicators require additional arguments. They should be added to print_indicators function with names used in ta-lib.

CODE EXAMPLE

from dataaggregation.aggregation import data_aggregation

my_data = data_aggregation('thelinkgoeshere')
my_data.get_data()
my_data.print_indicators('D', length = 14, smooth_step=6, indicators = ['RSI'])

Anomaly detection

Should only be used with a great understanding of features of data it was trained on. Works best on a volatile time series with high liquidity.

CODE EXAMPLE

from dataaggregation.aggregation import lstm_anomaly_detection

anms = lstm_anomaly_detection(url = 'thelinkgoeshere')
anms.detect_anomalies(threshold=1.25)

Output example####

NOW AVALIABLE ON PIP!

pip install python_storage_timeline_ta_indicators