The FreqGymNormalized strategy is a trading strategy that performs backtesting using various technical analysis (TA) indicators. Here's a short description of what the strategy does:
The populate_indicators method takes raw data from the exchange and adds several TA indicators to the DataFrame. These indicators include plus_di, minus_di, uo, htsine, htleadsine, bop, slowk, slowd, fastk, bb2_lower_gt_close, bb3_lower_gt_close, adx, aroonup, aroondown, aroonosc, cmo, dx, mfi, minus_di, plus_di, willr, rsi, fisher_rsi, stochrsi_k, stochrsi_d, and linangle.
The populate_buy_trend method generates the buy signal for the given DataFrame based on the TA indicators and an RL (reinforcement learning) model's prediction.
The populate_sell_trend method generates the sell signal for the given DataFrame based on the TA indicators and an RL model's prediction.
The strategy uses the RL model to predict the actions (buy/sell) based on the indicators. The buy and sell signals are added to the DataFrame as buy and sell columns, respectively. The strategy also includes some logging statements to track the progress and performance of indicator calculation and signal generation. This strategy aims to generate buy and sell signals based on the provided TA indicators and the RL model's predictions, allowing for backtesting and evaluation of the strategy's performance.