1.03.2023
- Remove bias from future tertiary data
18.02.2023
- Tertiary Patterns
- Correspond Tertiary pattern to upward or downward movement of market
08.02.2023
- change parametergen to generate exact after pattens aka not only [pattern]*^ but rather [pattern]^
- Test new parameters on 2019 EURUSD Data
- continues Search Saves all preview length libraries determine if they are usefully and if not prevent them from saving
- No powerful trend discovered max length: 10
03.02.2023
- sorted and cleaned package, better folder management
- Generate more broad pattern recognition
- change parametergen to generate exact after pattens aka not only [pattern]*^ but rather [pattern]^
02.02.2034
- load generated Trends in real.ipynb
- load Test Tick Data EURUSD 2020
- Detected Possible Flaw
- loading csv and converting to array will give string not float type. Comparison between string may be by length not by number.
- Convert Strings to floats
- Rerun trend finder with corrected floats
- Checked parametergen, working correctly
23.01.2023
- load generated Trends in real.ipynb and test for errors
14.01.2023
- Choose Tick and Period for UDpatterns recognition [ EURUSD whole 2019 year 1min ]
- Download Ticks and Save Locally [Data\EURUSD\DAT_MT_EURUSD_M1_2019.csv]
- Find Trends
- Choose alt_hypo_perc Threshold [0.05]
- Choose Trend Length Limit [10]
- Run trend_finder.py
- Count total trends found [8036]
- Check Everything
- find if there is reason why only trends with p < 0.36755 [till trend length 8 no trend with greater than 0.6 was detected - likely do to data itself]
- Test Trends
- Choose Ticks for testing [ EURUSD whole 2020 year 1min]
- Download Ticks and Save Locally
- load generated Trends in real.ipynb and test for errors