Zdrojový kód wiki Algotrading
Version 11.1 by David Brazda on 26.02.2023 14:45
Show last authors
| author | version | line-number | content |
|---|---|---|---|
| 1 | = Resources = | ||
| 2 | |||
| 3 | algo-trading basics [[https:~~/~~/algo-trading.readthedocs.io/en/latest/data-science-basics.html>>https://algo-trading.readthedocs.io/en/latest/data-science-basics.html]] a [[github>>https://github.com/awoo424/algotrading]] | ||
| 4 | |||
| 5 | |||
| 6 | = Python and stocks = | ||
| 7 | |||
| 8 | [[https:~~/~~/github.com/LastAncientOne/SimpleStockAnalysisPython>>https://github.com/LastAncientOne/SimpleStockAnalysisPython]] | ||
| 9 | |||
| 10 | [[integrace s TradingView webhooks >>https://github.com/robswc/tradingview-webhooks-bot]]- python based, na vlastnim serveru. (pokud se osvedci traderspost, pak muzu nainstalovat toto, napriklad na forum.pgwiki.cz, kde je docker a ma domenu) | ||
| 11 | |||
| 12 | [[python AI - pandas market predictor on GITHUB>>https://github.com/somkietacode/Pandas_Market_Predictor]] | ||
| 13 | |||
| 14 | = Zajimavé indikátory = | ||
| 15 | |||
| 16 | Podívat se na Prior SAR: The SAR value for previous period. | ||
| 17 | Extreme Point (EP): The highest high of the current uptrend. | ||
| 18 | Acceleration Factor (AF): Starting at 0.02, increases by 0.02 each time the extreme point makes a new high. AF can only reach a maximum of 0.2, no matter how long the uptrend extends. | ||
| 19 | \\Dále se podívat na Commodity Channel Index (CCI) | ||
| 20 | The Commodity Channel Index (CCI) helps identify price reversals, price extremes, and trend strength. | ||
| 21 | Developed by Donald Lambert, CCI is a momentum-based oscillator used to help determine when an investment vehicle is reaching a condition of being overbought or oversold. Popis zde - vyzkouset | ||
| 22 | |||
| 23 | A také [[ROC>>https://algo-trading.readthedocs.io/en/latest/technical-analysis.html#rate-of-change-roc]] | ||
| 24 | |||
| 25 | Popis základních indikátorů a jak na ně v pythonu je zde (projít!!) | ||
| 26 | [[https:~~/~~/algo-trading.readthedocs.io/en/latest/technical-analysis.html>>url:https://algo-trading.readthedocs.io/en/latest/technical-analysis.html]] | ||
| 27 | \\a dale | ||
| 28 | Rsi divergence | ||
| 29 | [[https:~~/~~/github.com/SpiralDevelopment/RSI-divergence-detector>>url:https://github.com/SpiralDevelopment/RSI-divergence-detector]] | ||
| 30 | |||
| 31 | |||
| 32 | = AI = | ||
| 33 | |||
| 34 | Základní myšlenka: Vstup (buď trade data nebo nějaký sec rt bar) + základní hodnoty indikátorů(ma, rsi, momentum..). Model dostane sušenku, když v aktuálním bodě predikuje, že půjde o malou hodnotu nahoru (např. 5 ticků). Za to dostane odměnu. A pustit na to třeba měsíční data. A vidělo by se. To by nemuselo být složité na interpretaci. | ||
| 35 | |||
| 36 | A pak vytrénovaný model pustit na online data. | ||
| 37 | |||
| 38 | |||
| 39 | LSTM in stock - příklad [[https:~~/~~/www.kaggle.com/code/faressayah/stock-market-analysis-prediction-using-lstm/notebook>>https://www.kaggle.com/code/faressayah/stock-market-analysis-prediction-using-lstm/notebook]] | ||
| 40 | |||
| 41 | podívat se na keras a LTSM | ||
| 42 | |||
| 43 | Zajimavý článek [[základy Recurrent Neural Networks>>http://karpathy.github.io/2015/05/21/rnn-effectiveness/]] a [[https:~~/~~/towardsdatascience.com/recurrent-neural-networks-rnns-3f06d7653a85>>https://towardsdatascience.com/recurrent-neural-networks-rnns-3f06d7653a85]] | ||
| 44 | |||
| 45 | Příklad - [[singlefeature LSTM model in pytorch>>https://algo-trading.readthedocs.io/en/latest/integrated-strategy.html#single-feature-lstm-model]]. | ||
| 46 | |||
| 47 | [[Stock Machine learning algorithms on github>>https://github.com/LastAncientOne/Deep_Learning_Machine_Learning_Stock]] ([[LSTM>>https://github.com/LastAncientOne/Deep_Learning_Machine_Learning_Stock/blob/master/Stock_Algorithms/LSTM_RNN_Part2.ipynb]])- projít | ||
| 48 | |||
| 49 | Zadání: predikční modely (HFT, 15s dt, střednědobý) | ||
| 50 | |||
| 51 | * - learning data | ||
| 52 | ** vývoj ceny - timeseries relevantní frekvence (trades, 15s ohlc a třeba 1h(1d) ohlc) | ||
| 53 | ** dopočítané/doplněné sloupce k timeseries (indikátory, calendar data a další time related data) | ||
| 54 | ** další doplňující atributy - jako premarket, fair price, trend z minulých období, odhad dnešního dne(dopočítávaný), celkový sentiment(může doplňovat jiný [[model>>https://algo-trading.readthedocs.io/en/latest/sentiment-analysis.html]] z news) | ||
| 55 | * | ||
| 56 | |||
| 57 |