From c6c569b77288519c5a7d2016d6a291ee8d68b9ea Mon Sep 17 00:00:00 2001 From: Matthias Date: Sat, 30 Apr 2022 14:47:27 +0200 Subject: [PATCH] chore: split BTAnalyais to metrics --- freqtrade/data/btanalysis.py | 167 +---------------- freqtrade/data/metrics.py | 173 ++++++++++++++++++ .../hyperopt_loss/hyperopt_loss_calmar.py | 2 +- .../hyperopt_loss_max_drawdown.py | 2 +- .../hyperopt_loss_profit_drawdown.py | 2 +- freqtrade/optimize/optimize_reports.py | 4 +- freqtrade/plot/plotting.py | 7 +- .../protections/max_drawdown_protection.py | 2 +- tests/data/test_btanalysis.py | 8 +- tests/test_plotting.py | 3 +- 10 files changed, 190 insertions(+), 180 deletions(-) create mode 100644 freqtrade/data/metrics.py diff --git a/freqtrade/data/btanalysis.py b/freqtrade/data/btanalysis.py index 0c8e721c0..e29d9ebe4 100644 --- a/freqtrade/data/btanalysis.py +++ b/freqtrade/data/btanalysis.py @@ -5,7 +5,7 @@ import logging from copy import copy from datetime import datetime, timezone from pathlib import Path -from typing import Any, Dict, List, Optional, Tuple, Union +from typing import Any, Dict, List, Optional, Union import numpy as np import pandas as pd @@ -400,168 +400,3 @@ def extract_trades_of_period(dataframe: pd.DataFrame, trades: pd.DataFrame, trades = trades.loc[(trades['open_date'] >= trades_start) & (trades['close_date'] <= trades_stop)] return trades - - -def calculate_market_change(data: Dict[str, pd.DataFrame], column: str = "close") -> float: - """ - Calculate market change based on "column". - Calculation is done by taking the first non-null and the last non-null element of each column - and calculating the pctchange as "(last - first) / first". - Then the results per pair are combined as mean. - - :param data: Dict of Dataframes, dict key should be pair. - :param column: Column in the original dataframes to use - :return: - """ - tmp_means = [] - for pair, df in data.items(): - start = df[column].dropna().iloc[0] - end = df[column].dropna().iloc[-1] - tmp_means.append((end - start) / start) - - return float(np.mean(tmp_means)) - - -def combine_dataframes_with_mean(data: Dict[str, pd.DataFrame], - column: str = "close") -> pd.DataFrame: - """ - Combine multiple dataframes "column" - :param data: Dict of Dataframes, dict key should be pair. - :param column: Column in the original dataframes to use - :return: DataFrame with the column renamed to the dict key, and a column - named mean, containing the mean of all pairs. - :raise: ValueError if no data is provided. - """ - df_comb = pd.concat([data[pair].set_index('date').rename( - {column: pair}, axis=1)[pair] for pair in data], axis=1) - - df_comb['mean'] = df_comb.mean(axis=1) - - return df_comb - - -def create_cum_profit(df: pd.DataFrame, trades: pd.DataFrame, col_name: str, - timeframe: str) -> pd.DataFrame: - """ - Adds a column `col_name` with the cumulative profit for the given trades array. - :param df: DataFrame with date index - :param trades: DataFrame containing trades (requires columns close_date and profit_abs) - :param col_name: Column name that will be assigned the results - :param timeframe: Timeframe used during the operations - :return: Returns df with one additional column, col_name, containing the cumulative profit. - :raise: ValueError if trade-dataframe was found empty. - """ - if len(trades) == 0: - raise ValueError("Trade dataframe empty.") - from freqtrade.exchange import timeframe_to_minutes - timeframe_minutes = timeframe_to_minutes(timeframe) - # Resample to timeframe to make sure trades match candles - _trades_sum = trades.resample(f'{timeframe_minutes}min', on='close_date' - )[['profit_abs']].sum() - df.loc[:, col_name] = _trades_sum['profit_abs'].cumsum() - # Set first value to 0 - df.loc[df.iloc[0].name, col_name] = 0 - # FFill to get continuous - df[col_name] = df[col_name].ffill() - return df - - -def _calc_drawdown_series(profit_results: pd.DataFrame, *, date_col: str, value_col: str - ) -> pd.DataFrame: - max_drawdown_df = pd.DataFrame() - max_drawdown_df['cumulative'] = profit_results[value_col].cumsum() - max_drawdown_df['high_value'] = max_drawdown_df['cumulative'].cummax() - max_drawdown_df['drawdown'] = max_drawdown_df['cumulative'] - max_drawdown_df['high_value'] - max_drawdown_df['date'] = profit_results.loc[:, date_col] - return max_drawdown_df - - -def calculate_underwater(trades: pd.DataFrame, *, date_col: str = 'close_date', - value_col: str = 'profit_ratio' - ): - """ - Calculate max drawdown and the corresponding close dates - :param trades: DataFrame containing trades (requires columns close_date and profit_ratio) - :param date_col: Column in DataFrame to use for dates (defaults to 'close_date') - :param value_col: Column in DataFrame to use for values (defaults to 'profit_ratio') - :return: Tuple (float, highdate, lowdate, highvalue, lowvalue) with absolute max drawdown, - high and low time and high and low value. - :raise: ValueError if trade-dataframe was found empty. - """ - if len(trades) == 0: - raise ValueError("Trade dataframe empty.") - profit_results = trades.sort_values(date_col).reset_index(drop=True) - max_drawdown_df = _calc_drawdown_series(profit_results, date_col=date_col, value_col=value_col) - - return max_drawdown_df - - -def calculate_max_drawdown(trades: pd.DataFrame, *, date_col: str = 'close_date', - value_col: str = 'profit_abs', starting_balance: float = 0 - ) -> Tuple[float, pd.Timestamp, pd.Timestamp, float, float, float]: - """ - Calculate max drawdown and the corresponding close dates - :param trades: DataFrame containing trades (requires columns close_date and profit_ratio) - :param date_col: Column in DataFrame to use for dates (defaults to 'close_date') - :param value_col: Column in DataFrame to use for values (defaults to 'profit_abs') - :param starting_balance: Portfolio starting balance - properly calculate relative drawdown. - :return: Tuple (float, highdate, lowdate, highvalue, lowvalue, relative_drawdown) - with absolute max drawdown, high and low time and high and low value, - and the relative account drawdown - :raise: ValueError if trade-dataframe was found empty. - """ - if len(trades) == 0: - raise ValueError("Trade dataframe empty.") - profit_results = trades.sort_values(date_col).reset_index(drop=True) - max_drawdown_df = _calc_drawdown_series(profit_results, date_col=date_col, value_col=value_col) - - idxmin = max_drawdown_df['drawdown'].idxmin() - if idxmin == 0: - raise ValueError("No losing trade, therefore no drawdown.") - high_date = profit_results.loc[max_drawdown_df.iloc[:idxmin]['high_value'].idxmax(), date_col] - low_date = profit_results.loc[idxmin, date_col] - high_val = max_drawdown_df.loc[max_drawdown_df.iloc[:idxmin] - ['high_value'].idxmax(), 'cumulative'] - low_val = max_drawdown_df.loc[idxmin, 'cumulative'] - max_drawdown_rel = 0.0 - if high_val + starting_balance != 0: - max_drawdown_rel = (high_val - low_val) / (high_val + starting_balance) - - return ( - abs(min(max_drawdown_df['drawdown'])), - high_date, - low_date, - high_val, - low_val, - max_drawdown_rel - ) - - -def calculate_csum(trades: pd.DataFrame, starting_balance: float = 0) -> Tuple[float, float]: - """ - Calculate min/max cumsum of trades, to show if the wallet/stake amount ratio is sane - :param trades: DataFrame containing trades (requires columns close_date and profit_percent) - :param starting_balance: Add starting balance to results, to show the wallets high / low points - :return: Tuple (float, float) with cumsum of profit_abs - :raise: ValueError if trade-dataframe was found empty. - """ - if len(trades) == 0: - raise ValueError("Trade dataframe empty.") - - csum_df = pd.DataFrame() - csum_df['sum'] = trades['profit_abs'].cumsum() - csum_min = csum_df['sum'].min() + starting_balance - csum_max = csum_df['sum'].max() + starting_balance - - return csum_min, csum_max - - -def calculate_cagr(days_passed: int, starting_balance: float, final_balance: float) -> float: - """ - Calculate CAGR - :param days_passed: Days passed between start and ending balance - :param starting_balance: Starting balance - :param final_balance: Final balance to calculate CAGR against - :return: CAGR - """ - return (final_balance / starting_balance) ** (1 / (days_passed / 365)) - 1 diff --git a/freqtrade/data/metrics.py b/freqtrade/data/metrics.py new file mode 100644 index 000000000..44d5ce6ec --- /dev/null +++ b/freqtrade/data/metrics.py @@ -0,0 +1,173 @@ +import logging +from typing import Dict, Tuple + +import numpy as np +import pandas as pd + + +logger = logging.getLogger(__name__) + + +def calculate_market_change(data: Dict[str, pd.DataFrame], column: str = "close") -> float: + """ + Calculate market change based on "column". + Calculation is done by taking the first non-null and the last non-null element of each column + and calculating the pctchange as "(last - first) / first". + Then the results per pair are combined as mean. + + :param data: Dict of Dataframes, dict key should be pair. + :param column: Column in the original dataframes to use + :return: + """ + tmp_means = [] + for pair, df in data.items(): + start = df[column].dropna().iloc[0] + end = df[column].dropna().iloc[-1] + tmp_means.append((end - start) / start) + + return float(np.mean(tmp_means)) + + +def combine_dataframes_with_mean(data: Dict[str, pd.DataFrame], + column: str = "close") -> pd.DataFrame: + """ + Combine multiple dataframes "column" + :param data: Dict of Dataframes, dict key should be pair. + :param column: Column in the original dataframes to use + :return: DataFrame with the column renamed to the dict key, and a column + named mean, containing the mean of all pairs. + :raise: ValueError if no data is provided. + """ + df_comb = pd.concat([data[pair].set_index('date').rename( + {column: pair}, axis=1)[pair] for pair in data], axis=1) + + df_comb['mean'] = df_comb.mean(axis=1) + + return df_comb + + +def create_cum_profit(df: pd.DataFrame, trades: pd.DataFrame, col_name: str, + timeframe: str) -> pd.DataFrame: + """ + Adds a column `col_name` with the cumulative profit for the given trades array. + :param df: DataFrame with date index + :param trades: DataFrame containing trades (requires columns close_date and profit_abs) + :param col_name: Column name that will be assigned the results + :param timeframe: Timeframe used during the operations + :return: Returns df with one additional column, col_name, containing the cumulative profit. + :raise: ValueError if trade-dataframe was found empty. + """ + if len(trades) == 0: + raise ValueError("Trade dataframe empty.") + from freqtrade.exchange import timeframe_to_minutes + timeframe_minutes = timeframe_to_minutes(timeframe) + # Resample to timeframe to make sure trades match candles + _trades_sum = trades.resample(f'{timeframe_minutes}min', on='close_date' + )[['profit_abs']].sum() + df.loc[:, col_name] = _trades_sum['profit_abs'].cumsum() + # Set first value to 0 + df.loc[df.iloc[0].name, col_name] = 0 + # FFill to get continuous + df[col_name] = df[col_name].ffill() + return df + + +def _calc_drawdown_series(profit_results: pd.DataFrame, *, date_col: str, value_col: str + ) -> pd.DataFrame: + max_drawdown_df = pd.DataFrame() + max_drawdown_df['cumulative'] = profit_results[value_col].cumsum() + max_drawdown_df['high_value'] = max_drawdown_df['cumulative'].cummax() + max_drawdown_df['drawdown'] = max_drawdown_df['cumulative'] - max_drawdown_df['high_value'] + max_drawdown_df['date'] = profit_results.loc[:, date_col] + return max_drawdown_df + + +def calculate_underwater(trades: pd.DataFrame, *, date_col: str = 'close_date', + value_col: str = 'profit_ratio' + ): + """ + Calculate max drawdown and the corresponding close dates + :param trades: DataFrame containing trades (requires columns close_date and profit_ratio) + :param date_col: Column in DataFrame to use for dates (defaults to 'close_date') + :param value_col: Column in DataFrame to use for values (defaults to 'profit_ratio') + :return: Tuple (float, highdate, lowdate, highvalue, lowvalue) with absolute max drawdown, + high and low time and high and low value. + :raise: ValueError if trade-dataframe was found empty. + """ + if len(trades) == 0: + raise ValueError("Trade dataframe empty.") + profit_results = trades.sort_values(date_col).reset_index(drop=True) + max_drawdown_df = _calc_drawdown_series(profit_results, date_col=date_col, value_col=value_col) + + return max_drawdown_df + + +def calculate_max_drawdown(trades: pd.DataFrame, *, date_col: str = 'close_date', + value_col: str = 'profit_abs', starting_balance: float = 0 + ) -> Tuple[float, pd.Timestamp, pd.Timestamp, float, float, float]: + """ + Calculate max drawdown and the corresponding close dates + :param trades: DataFrame containing trades (requires columns close_date and profit_ratio) + :param date_col: Column in DataFrame to use for dates (defaults to 'close_date') + :param value_col: Column in DataFrame to use for values (defaults to 'profit_abs') + :param starting_balance: Portfolio starting balance - properly calculate relative drawdown. + :return: Tuple (float, highdate, lowdate, highvalue, lowvalue, relative_drawdown) + with absolute max drawdown, high and low time and high and low value, + and the relative account drawdown + :raise: ValueError if trade-dataframe was found empty. + """ + if len(trades) == 0: + raise ValueError("Trade dataframe empty.") + profit_results = trades.sort_values(date_col).reset_index(drop=True) + max_drawdown_df = _calc_drawdown_series(profit_results, date_col=date_col, value_col=value_col) + + idxmin = max_drawdown_df['drawdown'].idxmin() + if idxmin == 0: + raise ValueError("No losing trade, therefore no drawdown.") + high_date = profit_results.loc[max_drawdown_df.iloc[:idxmin]['high_value'].idxmax(), date_col] + low_date = profit_results.loc[idxmin, date_col] + high_val = max_drawdown_df.loc[max_drawdown_df.iloc[:idxmin] + ['high_value'].idxmax(), 'cumulative'] + low_val = max_drawdown_df.loc[idxmin, 'cumulative'] + max_drawdown_rel = 0.0 + if high_val + starting_balance != 0: + max_drawdown_rel = (high_val - low_val) / (high_val + starting_balance) + + return ( + abs(min(max_drawdown_df['drawdown'])), + high_date, + low_date, + high_val, + low_val, + max_drawdown_rel + ) + + +def calculate_csum(trades: pd.DataFrame, starting_balance: float = 0) -> Tuple[float, float]: + """ + Calculate min/max cumsum of trades, to show if the wallet/stake amount ratio is sane + :param trades: DataFrame containing trades (requires columns close_date and profit_percent) + :param starting_balance: Add starting balance to results, to show the wallets high / low points + :return: Tuple (float, float) with cumsum of profit_abs + :raise: ValueError if trade-dataframe was found empty. + """ + if len(trades) == 0: + raise ValueError("Trade dataframe empty.") + + csum_df = pd.DataFrame() + csum_df['sum'] = trades['profit_abs'].cumsum() + csum_min = csum_df['sum'].min() + starting_balance + csum_max = csum_df['sum'].max() + starting_balance + + return csum_min, csum_max + + +def calculate_cagr(days_passed: int, starting_balance: float, final_balance: float) -> float: + """ + Calculate CAGR + :param days_passed: Days passed between start and ending balance + :param starting_balance: Starting balance + :param final_balance: Final balance to calculate CAGR against + :return: CAGR + """ + return (final_balance / starting_balance) ** (1 / (days_passed / 365)) - 1 diff --git a/freqtrade/optimize/hyperopt_loss/hyperopt_loss_calmar.py b/freqtrade/optimize/hyperopt_loss/hyperopt_loss_calmar.py index 846dae9ea..ea6c151e5 100644 --- a/freqtrade/optimize/hyperopt_loss/hyperopt_loss_calmar.py +++ b/freqtrade/optimize/hyperopt_loss/hyperopt_loss_calmar.py @@ -10,7 +10,7 @@ from typing import Any, Dict from pandas import DataFrame -from freqtrade.data.btanalysis import calculate_max_drawdown +from freqtrade.data.metrics import calculate_max_drawdown from freqtrade.optimize.hyperopt import IHyperOptLoss diff --git a/freqtrade/optimize/hyperopt_loss/hyperopt_loss_max_drawdown.py b/freqtrade/optimize/hyperopt_loss/hyperopt_loss_max_drawdown.py index ce955d928..a8af704cd 100644 --- a/freqtrade/optimize/hyperopt_loss/hyperopt_loss_max_drawdown.py +++ b/freqtrade/optimize/hyperopt_loss/hyperopt_loss_max_drawdown.py @@ -8,7 +8,7 @@ from datetime import datetime from pandas import DataFrame -from freqtrade.data.btanalysis import calculate_max_drawdown +from freqtrade.data.metrics import calculate_max_drawdown from freqtrade.optimize.hyperopt import IHyperOptLoss diff --git a/freqtrade/optimize/hyperopt_loss/hyperopt_loss_profit_drawdown.py b/freqtrade/optimize/hyperopt_loss/hyperopt_loss_profit_drawdown.py index 5bd12ff52..ed689edba 100644 --- a/freqtrade/optimize/hyperopt_loss/hyperopt_loss_profit_drawdown.py +++ b/freqtrade/optimize/hyperopt_loss/hyperopt_loss_profit_drawdown.py @@ -9,7 +9,7 @@ individual needs. """ from pandas import DataFrame -from freqtrade.data.btanalysis import calculate_max_drawdown +from freqtrade.data.metrics import calculate_max_drawdown from freqtrade.optimize.hyperopt import IHyperOptLoss diff --git a/freqtrade/optimize/optimize_reports.py b/freqtrade/optimize/optimize_reports.py index 1d58dc339..9c1a276a9 100644 --- a/freqtrade/optimize/optimize_reports.py +++ b/freqtrade/optimize/optimize_reports.py @@ -9,8 +9,8 @@ from pandas import DataFrame, to_datetime from tabulate import tabulate from freqtrade.constants import DATETIME_PRINT_FORMAT, LAST_BT_RESULT_FN, UNLIMITED_STAKE_AMOUNT -from freqtrade.data.btanalysis import (calculate_cagr, calculate_csum, calculate_market_change, - calculate_max_drawdown) +from freqtrade.data.metrics import (calculate_cagr, calculate_csum, calculate_market_change, + calculate_max_drawdown) from freqtrade.misc import decimals_per_coin, file_dump_joblib, file_dump_json, round_coin_value from freqtrade.optimize.backtest_caching import get_backtest_metadata_filename diff --git a/freqtrade/plot/plotting.py b/freqtrade/plot/plotting.py index 5337016f3..773577d7b 100644 --- a/freqtrade/plot/plotting.py +++ b/freqtrade/plot/plotting.py @@ -5,12 +5,13 @@ from typing import Any, Dict, List, Optional import pandas as pd from freqtrade.configuration import TimeRange -from freqtrade.data.btanalysis import (analyze_trade_parallelism, calculate_max_drawdown, - calculate_underwater, combine_dataframes_with_mean, - create_cum_profit, extract_trades_of_period, load_trades) +from freqtrade.data.btanalysis import (analyze_trade_parallelism, extract_trades_of_period, + load_trades) from freqtrade.data.converter import trim_dataframe from freqtrade.data.dataprovider import DataProvider from freqtrade.data.history import get_timerange, load_data +from freqtrade.data.metrics import (calculate_max_drawdown, calculate_underwater, + combine_dataframes_with_mean, create_cum_profit) from freqtrade.enums import CandleType from freqtrade.exceptions import OperationalException from freqtrade.exchange import timeframe_to_prev_date, timeframe_to_seconds diff --git a/freqtrade/plugins/protections/max_drawdown_protection.py b/freqtrade/plugins/protections/max_drawdown_protection.py index b6ef92bd5..4111b7ff4 100644 --- a/freqtrade/plugins/protections/max_drawdown_protection.py +++ b/freqtrade/plugins/protections/max_drawdown_protection.py @@ -5,7 +5,7 @@ from typing import Any, Dict import pandas as pd -from freqtrade.data.btanalysis import calculate_max_drawdown +from freqtrade.data.metrics import calculate_max_drawdown from freqtrade.persistence import Trade from freqtrade.plugins.protections import IProtection, ProtectionReturn diff --git a/tests/data/test_btanalysis.py b/tests/data/test_btanalysis.py index eaf703b2d..f9f49e280 100644 --- a/tests/data/test_btanalysis.py +++ b/tests/data/test_btanalysis.py @@ -8,14 +8,14 @@ from pandas import DataFrame, DateOffset, Timestamp, to_datetime from freqtrade.configuration import TimeRange from freqtrade.constants import LAST_BT_RESULT_FN -from freqtrade.data.btanalysis import (BT_DATA_COLUMNS, analyze_trade_parallelism, calculate_cagr, - calculate_csum, calculate_market_change, - calculate_max_drawdown, calculate_underwater, - combine_dataframes_with_mean, create_cum_profit, +from freqtrade.data.btanalysis import (BT_DATA_COLUMNS, analyze_trade_parallelism, extract_trades_of_period, get_latest_backtest_filename, get_latest_hyperopt_file, load_backtest_data, load_backtest_metadata, load_trades, load_trades_from_db) from freqtrade.data.history import load_data, load_pair_history +from freqtrade.data.metrics import (calculate_cagr, calculate_csum, calculate_market_change, + calculate_max_drawdown, calculate_underwater, + combine_dataframes_with_mean, create_cum_profit) from freqtrade.exceptions import OperationalException from tests.conftest import CURRENT_TEST_STRATEGY, create_mock_trades from tests.conftest_trades import MOCK_TRADE_COUNT diff --git a/tests/test_plotting.py b/tests/test_plotting.py index 97f367608..65df2d84c 100644 --- a/tests/test_plotting.py +++ b/tests/test_plotting.py @@ -10,7 +10,8 @@ from plotly.subplots import make_subplots from freqtrade.commands import start_plot_dataframe, start_plot_profit from freqtrade.configuration import TimeRange from freqtrade.data import history -from freqtrade.data.btanalysis import create_cum_profit, load_backtest_data +from freqtrade.data.btanalysis import load_backtest_data +from freqtrade.data.metrics import create_cum_profit from freqtrade.exceptions import OperationalException from freqtrade.plot.plotting import (add_areas, add_indicators, add_profit, create_plotconfig, generate_candlestick_graph, generate_plot_filename,