代码拉取完成，页面将自动刷新
同步操作将从 云金杞/pyfolio 强制同步，此操作会覆盖自 Fork 仓库以来所做的任何修改，且无法恢复！！！
确定后同步将在后台操作，完成时将刷新页面，请耐心等待。
#
# Copyright 2018 Quantopian, Inc.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
from __future__ import division
from collections import OrderedDict
from functools import partial
import empyrical as ep
import numpy as np
import pandas as pd
import scipy as sp
import scipy.stats as stats
from .deprecate import deprecated
from .interesting_periods import PERIODS
from .txn import get_turnover
from .utils import APPROX_BDAYS_PER_MONTH, APPROX_BDAYS_PER_YEAR
from .utils import DAILY
DEPRECATION_WARNING = ("Risk functions in pyfolio.timeseries are deprecated "
"and will be removed in a future release. Please "
"install the empyrical package instead.")
def var_cov_var_normal(P, c, mu=0, sigma=1):
"""
Variance-covariance calculation of daily Value-at-Risk in a
portfolio.
Parameters
----------
P : float
Portfolio value.
c : float
Confidence level.
mu : float, optional
Mean.
Returns
-------
float
Variance-covariance.
"""
alpha = sp.stats.norm.ppf(1 - c, mu, sigma)
return P - P * (alpha + 1)
@deprecated(msg=DEPRECATION_WARNING)
def max_drawdown(returns):
"""
Determines the maximum drawdown of a strategy.
Parameters
----------
returns : pd.Series
Daily returns of the strategy, noncumulative.
- See full explanation in tears.create_full_tear_sheet.
Returns
-------
float
Maximum drawdown.
Note
-----
See https://en.wikipedia.org/wiki/Drawdown_(economics) for more details.
"""
return ep.max_drawdown(returns)
@deprecated(msg=DEPRECATION_WARNING)
def annual_return(returns, period=DAILY):
"""
Determines the mean annual growth rate of returns.
Parameters
----------
returns : pd.Series
Periodic returns of the strategy, noncumulative.
- See full explanation in :func:`~pyfolio.timeseries.cum_returns`.
period : str, optional
Defines the periodicity of the 'returns' data for purposes of
annualizing. Can be 'monthly', 'weekly', or 'daily'.
- Defaults to 'daily'.
Returns
-------
float
Annual Return as CAGR (Compounded Annual Growth Rate).
"""
return ep.annual_return(returns, period=period)
@deprecated(msg=DEPRECATION_WARNING)
def annual_volatility(returns, period=DAILY):
"""
Determines the annual volatility of a strategy.
Parameters
----------
returns : pd.Series
Periodic returns of the strategy, noncumulative.
- See full explanation in :func:`~pyfolio.timeseries.cum_returns`.
period : str, optional
Defines the periodicity of the 'returns' data for purposes of
annualizing volatility. Can be 'monthly' or 'weekly' or 'daily'.
- Defaults to 'daily'.
Returns
-------
float
Annual volatility.
"""
return ep.annual_volatility(returns, period=period)
@deprecated(msg=DEPRECATION_WARNING)
def calmar_ratio(returns, period=DAILY):
"""
Determines the Calmar ratio, or drawdown ratio, of a strategy.
Parameters
----------
returns : pd.Series
Daily returns of the strategy, noncumulative.
- See full explanation in :func:`~pyfolio.timeseries.cum_returns`.
period : str, optional
Defines the periodicity of the 'returns' data for purposes of
annualizing. Can be 'monthly', 'weekly', or 'daily'.
- Defaults to 'daily'.
Returns
-------
float
Calmar ratio (drawdown ratio) as float. Returns np.nan if there is no
calmar ratio.
Note
-----
See https://en.wikipedia.org/wiki/Calmar_ratio for more details.
"""
return ep.calmar_ratio(returns, period=period)
@deprecated(msg=DEPRECATION_WARNING)
def omega_ratio(returns, annual_return_threshhold=0.0):
"""
Determines the Omega ratio of a strategy.
Parameters
----------
returns : pd.Series
Daily returns of the strategy, noncumulative.
- See full explanation in :func:`~pyfolio.timeseries.cum_returns`.
annual_return_threshold : float, optional
Minimum acceptable return of the investor. Annual threshold over which
returns are considered positive or negative. It is converted to a
value appropriate for the period of the returns for this ratio.
E.g. An annual minimum acceptable return of 100 translates to a daily
minimum acceptable return of 0.01848.
(1 + 100) ** (1. / 252) - 1 = 0.01848
Daily returns must exceed this value to be considered positive. The
daily return yields the desired annual return when compounded over
the average number of business days in a year.
(1 + 0.01848) ** 252 - 1 = 99.93
- Defaults to 0.0
Returns
-------
float
Omega ratio.
Note
-----
See https://en.wikipedia.org/wiki/Omega_ratio for more details.
"""
return ep.omega_ratio(returns,
required_return=annual_return_threshhold)
@deprecated(msg=DEPRECATION_WARNING)
def sortino_ratio(returns, required_return=0, period=DAILY):
"""
Determines the Sortino ratio of a strategy.
Parameters
----------
returns : pd.Series or pd.DataFrame
Daily returns of the strategy, noncumulative.
- See full explanation in :func:`~pyfolio.timeseries.cum_returns`.
required_return: float / series
minimum acceptable return
period : str, optional
Defines the periodicity of the 'returns' data for purposes of
annualizing. Can be 'monthly', 'weekly', or 'daily'.
- Defaults to 'daily'.
Returns
-------
depends on input type
series ==> float
DataFrame ==> np.array
Annualized Sortino ratio.
"""
return ep.sortino_ratio(returns, required_return=required_return)
@deprecated(msg=DEPRECATION_WARNING)
def downside_risk(returns, required_return=0, period=DAILY):
"""
Determines the downside deviation below a threshold
Parameters
----------
returns : pd.Series or pd.DataFrame
Daily returns of the strategy, noncumulative.
- See full explanation in :func:`~pyfolio.timeseries.cum_returns`.
required_return: float / series
minimum acceptable return
period : str, optional
Defines the periodicity of the 'returns' data for purposes of
annualizing. Can be 'monthly', 'weekly', or 'daily'.
- Defaults to 'daily'.
Returns
-------
depends on input type
series ==> float
DataFrame ==> np.array
Annualized downside deviation
"""
return ep.downside_risk(returns,
required_return=required_return,
period=period)
@deprecated(msg=DEPRECATION_WARNING)
def sharpe_ratio(returns, risk_free=0, period=DAILY):
"""
Determines the Sharpe ratio of a strategy.
Parameters
----------
returns : pd.Series
Daily returns of the strategy, noncumulative.
- See full explanation in :func:`~pyfolio.timeseries.cum_returns`.
risk_free : int, float
Constant risk-free return throughout the period.
period : str, optional
Defines the periodicity of the 'returns' data for purposes of
annualizing. Can be 'monthly', 'weekly', or 'daily'.
- Defaults to 'daily'.
Returns
-------
float
Sharpe ratio.
np.nan
If insufficient length of returns or if if adjusted returns are 0.
Note
-----
See https://en.wikipedia.org/wiki/Sharpe_ratio for more details.
"""
return ep.sharpe_ratio(returns, risk_free=risk_free, period=period)
@deprecated(msg=DEPRECATION_WARNING)
def alpha_beta(returns, factor_returns):
"""
Calculates both alpha and beta.
Parameters
----------
returns : pd.Series
Daily returns of the strategy, noncumulative.
- See full explanation in :func:`~pyfolio.timeseries.cum_returns`.
factor_returns : pd.Series
Daily noncumulative returns of the benchmark factor to which betas are
computed. Usually a benchmark such as market returns.
- This is in the same style as returns.
Returns
-------
float
Alpha.
float
Beta.
"""
return ep.alpha_beta(returns, factor_returns=factor_returns)
@deprecated(msg=DEPRECATION_WARNING)
def alpha(returns, factor_returns):
"""
Calculates annualized alpha.
Parameters
----------
returns : pd.Series
Daily returns of the strategy, noncumulative.
- See full explanation in :func:`~pyfolio.timeseries.cum_returns`.
factor_returns : pd.Series
Daily noncumulative returns of the benchmark factor to which betas are
computed. Usually a benchmark such as market returns.
- This is in the same style as returns.
Returns
-------
float
Alpha.
"""
return ep.alpha(returns, factor_returns=factor_returns)
@deprecated(msg=DEPRECATION_WARNING)
def beta(returns, factor_returns):
"""
Calculates beta.
Parameters
----------
returns : pd.Series
Daily returns of the strategy, noncumulative.
- See full explanation in :func:`~pyfolio.timeseries.cum_returns`.
factor_returns : pd.Series
Daily noncumulative returns of the benchmark factor to which betas are
computed. Usually a benchmark such as market returns.
- This is in the same style as returns.
Returns
-------
float
Beta.
"""
return ep.beta(returns, factor_returns)
@deprecated(msg=DEPRECATION_WARNING)
def stability_of_timeseries(returns):
"""
Determines R-squared of a linear fit to the cumulative
log returns. Computes an ordinary least squares linear fit,
and returns R-squared.
Parameters
----------
returns : pd.Series
Daily returns of the strategy, noncumulative.
- See full explanation in :func:`~pyfolio.timeseries.cum_returns`.
Returns
-------
float
R-squared.
"""
return ep.stability_of_timeseries(returns)
@deprecated(msg=DEPRECATION_WARNING)
def tail_ratio(returns):
"""
Determines the ratio between the right (95%) and left tail (5%).
For example, a ratio of 0.25 means that losses are four times
as bad as profits.
Parameters
----------
returns : pd.Series
Daily returns of the strategy, noncumulative.
- See full explanation in :func:`~pyfolio.timeseries.cum_returns`.
Returns
-------
float
tail ratio
"""
return ep.tail_ratio(returns)
def common_sense_ratio(returns):
"""
Common sense ratio is the multiplication of the tail ratio and the
Gain-to-Pain-Ratio -- sum(profits) / sum(losses).
See http://bit.ly/1ORzGBk for more information on motivation of
this metric.
Parameters
----------
returns : pd.Series
Daily returns of the strategy, noncumulative.
- See full explanation in tears.create_full_tear_sheet.
Returns
-------
float
common sense ratio
"""
return ep.tail_ratio(returns) * \
(1 + ep.annual_return(returns))
def normalize(returns, starting_value=1):
"""
Normalizes a returns timeseries based on the first value.
Parameters
----------
returns : pd.Series
Daily returns of the strategy, noncumulative.
- See full explanation in tears.create_full_tear_sheet.
starting_value : float, optional
The starting returns (default 1).
Returns
-------
pd.Series
Normalized returns.
"""
return starting_value * (returns / returns.iloc[0])
@deprecated(msg=DEPRECATION_WARNING)
def cum_returns(returns, starting_value=0):
"""
Compute cumulative returns from simple returns.
Parameters
----------
returns : pd.Series
Daily returns of the strategy, noncumulative.
- See full explanation in tears.create_full_tear_sheet.
starting_value : float, optional
The starting returns (default 1).
Returns
-------
pandas.Series
Series of cumulative returns.
Notes
-----
For increased numerical accuracy, convert input to log returns
where it is possible to sum instead of multiplying.
"""
return ep.cum_returns(returns, starting_value=starting_value)
@deprecated(msg=DEPRECATION_WARNING)
def aggregate_returns(returns, convert_to):
"""
Aggregates returns by week, month, or year.
Parameters
----------
returns : pd.Series
Daily returns of the strategy, noncumulative.
- See full explanation in :func:`~pyfolio.timeseries.cum_returns`.
convert_to : str
Can be 'weekly', 'monthly', or 'yearly'.
Returns
-------
pd.Series
Aggregated returns.
"""
return ep.aggregate_returns(returns, convert_to=convert_to)
def rolling_beta(returns, factor_returns,
rolling_window=APPROX_BDAYS_PER_MONTH * 6):
"""
Determines the rolling beta of a strategy.
Parameters
----------
returns : pd.Series
Daily returns of the strategy, noncumulative.
- See full explanation in tears.create_full_tear_sheet.
factor_returns : pd.Series or pd.DataFrame
Daily noncumulative returns of the benchmark factor to which betas are
computed. Usually a benchmark such as market returns.
- If DataFrame is passed, computes rolling beta for each column.
- This is in the same style as returns.
rolling_window : int, optional
The size of the rolling window, in days, over which to compute
beta (default 6 months).
Returns
-------
pd.Series
Rolling beta.
Note
-----
See https://en.wikipedia.org/wiki/Beta_(finance) for more details.
"""
if factor_returns.ndim > 1:
# Apply column-wise
return factor_returns.apply(partial(rolling_beta, returns),
rolling_window=rolling_window)
else:
out = pd.Series(index=returns.index)
for beg, end in zip(returns.index[0:-rolling_window],
returns.index[rolling_window:]):
out.loc[end] = ep.beta(
returns.loc[beg:end],
factor_returns.loc[beg:end])
return out
def rolling_regression(returns, factor_returns,
rolling_window=APPROX_BDAYS_PER_MONTH * 6,
nan_threshold=0.1):
"""
Computes rolling factor betas using a multivariate linear regression
(separate linear regressions is problematic because the factors may be
confounded).
Parameters
----------
returns : pd.Series
Daily returns of the strategy, noncumulative.
- See full explanation in tears.create_full_tear_sheet.
factor_returns : pd.DataFrame
Daily noncumulative returns of the benchmark factor to which betas are
computed. Usually a benchmark such as market returns.
- Computes rolling beta for each column.
- This is in the same style as returns.
rolling_window : int, optional
The days window over which to compute the beta. Defaults to 6 months.
nan_threshold : float, optional
If there are more than this fraction of NaNs, the rolling regression
for the given date will be skipped.
Returns
-------
pandas.DataFrame
DataFrame containing rolling beta coefficients to SMB, HML and UMD
"""
# We need to drop NaNs to regress
from sklearn import linear_model
ret_no_na = returns.dropna()
columns = ['alpha'] + factor_returns.columns.tolist()
rolling_risk = pd.DataFrame(columns=columns,
index=ret_no_na.index)
rolling_risk.index.name = 'dt'
for beg, end in zip(ret_no_na.index[:-rolling_window],
ret_no_na.index[rolling_window:]):
returns_period = ret_no_na[beg:end]
factor_returns_period = factor_returns.loc[returns_period.index]
if np.all(factor_returns_period.isnull().mean()) < nan_threshold:
factor_returns_period_dnan = factor_returns_period.dropna()
reg = linear_model.LinearRegression(fit_intercept=True).fit(
factor_returns_period_dnan,
returns_period.loc[factor_returns_period_dnan.index])
rolling_risk.loc[end, factor_returns.columns] = reg.coef_
rolling_risk.loc[end, 'alpha'] = reg.intercept_
return rolling_risk
def gross_lev(positions):
"""
Calculates the gross leverage of a strategy.
Parameters
----------
positions : pd.DataFrame
Daily net position values.
- See full explanation in tears.create_full_tear_sheet.
Returns
-------
pd.Series
Gross leverage.
"""
exposure = positions.drop('cash', axis=1).abs().sum(axis=1)
return exposure / positions.sum(axis=1)
def value_at_risk(returns, period=None, sigma=2.0):
"""
Get value at risk (VaR).
Parameters
----------
returns : pd.Series
Daily returns of the strategy, noncumulative.
- See full explanation in tears.create_full_tear_sheet.
period : str, optional
Period over which to calculate VaR. Set to 'weekly',
'monthly', or 'yearly', otherwise defaults to period of
returns (typically daily).
sigma : float, optional
Standard deviations of VaR, default 2.
"""
if period is not None:
returns_agg = ep.aggregate_returns(returns, period)
else:
returns_agg = returns.copy()
value_at_risk = returns_agg.mean() - sigma * returns_agg.std()
return value_at_risk
SIMPLE_STAT_FUNCS = [
ep.annual_return,
ep.cum_returns_final,
ep.annual_volatility,
ep.sharpe_ratio,
ep.calmar_ratio,
ep.stability_of_timeseries,
ep.max_drawdown,
ep.omega_ratio,
ep.sortino_ratio,
stats.skew,
stats.kurtosis,
ep.tail_ratio,
value_at_risk
]
FACTOR_STAT_FUNCS = [
ep.alpha,
ep.beta,
]
STAT_FUNC_NAMES = {
'annual_return': 'Annual return',
'cum_returns_final': 'Cumulative returns',
'annual_volatility': 'Annual volatility',
'sharpe_ratio': 'Sharpe ratio',
'calmar_ratio': 'Calmar ratio',
'stability_of_timeseries': 'Stability',
'max_drawdown': 'Max drawdown',
'omega_ratio': 'Omega ratio',
'sortino_ratio': 'Sortino ratio',
'skew': 'Skew',
'kurtosis': 'Kurtosis',
'tail_ratio': 'Tail ratio',
'common_sense_ratio': 'Common sense ratio',
'value_at_risk': 'Daily value at risk',
'alpha': 'Alpha',
'beta': 'Beta',
}
def perf_stats(returns, factor_returns=None, positions=None,
transactions=None, turnover_denom='AGB'):
"""
Calculates various performance metrics of a strategy, for use in
plotting.show_perf_stats.
Parameters
----------
returns : pd.Series
Daily returns of the strategy, noncumulative.
- See full explanation in tears.create_full_tear_sheet.
factor_returns : pd.Series, optional
Daily noncumulative returns of the benchmark factor to which betas are
computed. Usually a benchmark such as market returns.
- This is in the same style as returns.
- If None, do not compute alpha, beta, and information ratio.
positions : pd.DataFrame
Daily net position values.
- See full explanation in tears.create_full_tear_sheet.
transactions : pd.DataFrame
Prices and amounts of executed trades. One row per trade.
- See full explanation in tears.create_full_tear_sheet.
turnover_denom : str
Either AGB or portfolio_value, default AGB.
- See full explanation in txn.get_turnover.
Returns
-------
pd.Series
Performance metrics.
"""
stats = pd.Series(dtype=np.float64)
for stat_func in SIMPLE_STAT_FUNCS:
stats[STAT_FUNC_NAMES[stat_func.__name__]] = stat_func(returns)
if positions is not None:
stats['Gross leverage'] = gross_lev(positions).mean()
if transactions is not None:
stats['Daily turnover'] = get_turnover(positions,
transactions,
turnover_denom).mean()
if factor_returns is not None:
for stat_func in FACTOR_STAT_FUNCS:
res = stat_func(returns, factor_returns)
stats[STAT_FUNC_NAMES[stat_func.__name__]] = res
return stats
def perf_stats_bootstrap(returns, factor_returns=None, return_stats=True,
**kwargs):
"""Calculates various bootstrapped performance metrics of a strategy.
Parameters
----------
returns : pd.Series
Daily returns of the strategy, noncumulative.
- See full explanation in tears.create_full_tear_sheet.
factor_returns : pd.Series, optional
Daily noncumulative returns of the benchmark factor to which betas are
computed. Usually a benchmark such as market returns.
- This is in the same style as returns.
- If None, do not compute alpha, beta, and information ratio.
return_stats : boolean (optional)
If True, returns a DataFrame of mean, median, 5 and 95 percentiles
for each perf metric.
If False, returns a DataFrame with the bootstrap samples for
each perf metric.
Returns
-------
pd.DataFrame
if return_stats is True:
- Distributional statistics of bootstrapped sampling
distribution of performance metrics.
if return_stats is False:
- Bootstrap samples for each performance metric.
"""
bootstrap_values = OrderedDict()
for stat_func in SIMPLE_STAT_FUNCS:
stat_name = STAT_FUNC_NAMES[stat_func.__name__]
bootstrap_values[stat_name] = calc_bootstrap(stat_func,
returns)
if factor_returns is not None:
for stat_func in FACTOR_STAT_FUNCS:
stat_name = STAT_FUNC_NAMES[stat_func.__name__]
bootstrap_values[stat_name] = calc_bootstrap(
stat_func,
returns,
factor_returns=factor_returns)
bootstrap_values = pd.DataFrame(bootstrap_values)
if return_stats:
stats = bootstrap_values.apply(calc_distribution_stats)
return stats.T[['mean', 'median', '5%', '95%']]
else:
return bootstrap_values
def calc_bootstrap(func, returns, *args, **kwargs):
"""Performs a bootstrap analysis on a user-defined function returning
a summary statistic.
Parameters
----------
func : function
Function that either takes a single array (commonly returns)
or two arrays (commonly returns and factor returns) and
returns a single value (commonly a summary
statistic). Additional args and kwargs are passed as well.
returns : pd.Series
Daily returns of the strategy, noncumulative.
- See full explanation in tears.create_full_tear_sheet.
factor_returns : pd.Series, optional
Daily noncumulative returns of the benchmark factor to which betas are
computed. Usually a benchmark such as market returns.
- This is in the same style as returns.
n_samples : int, optional
Number of bootstrap samples to draw. Default is 1000.
Increasing this will lead to more stable / accurate estimates.
Returns
-------
numpy.ndarray
Bootstrapped sampling distribution of passed in func.
"""
n_samples = kwargs.pop('n_samples', 1000)
out = np.empty(n_samples)
factor_returns = kwargs.pop('factor_returns', None)
for i in range(n_samples):
idx = np.random.randint(len(returns), size=len(returns))
returns_i = returns.iloc[idx].reset_index(drop=True)
if factor_returns is not None:
factor_returns_i = factor_returns.iloc[idx].reset_index(drop=True)
out[i] = func(returns_i, factor_returns_i,
*args, **kwargs)
else:
out[i] = func(returns_i,
*args, **kwargs)
return out
def calc_distribution_stats(x):
"""Calculate various summary statistics of data.
Parameters
----------
x : numpy.ndarray or pandas.Series
Array to compute summary statistics for.
Returns
-------
pandas.Series
Series containing mean, median, std, as well as 5, 25, 75 and
95 percentiles of passed in values.
"""
return pd.Series({'mean': np.mean(x),
'median': np.median(x),
'std': np.std(x),
'5%': np.percentile(x, 5),
'25%': np.percentile(x, 25),
'75%': np.percentile(x, 75),
'95%': np.percentile(x, 95),
'IQR': np.subtract.reduce(
np.percentile(x, [75, 25])),
})
def get_max_drawdown_underwater(underwater):
"""
Determines peak, valley, and recovery dates given an 'underwater'
DataFrame.
An underwater DataFrame is a DataFrame that has precomputed
rolling drawdown.
Parameters
----------
underwater : pd.Series
Underwater returns (rolling drawdown) of a strategy.
Returns
-------
peak : datetime
The maximum drawdown's peak.
valley : datetime
The maximum drawdown's valley.
recovery : datetime
The maximum drawdown's recovery.
"""
# valley = np.argmin(underwater) # end of the period
# # print(valley)
# # Find first 0
# peak = underwater[:valley][underwater[:valley] == 0].index[-1]
# # Find last 0
# try:
# recovery = underwater[valley:][underwater[valley:] == 0].index[0]
# except IndexError:
# recovery = np.nan # drawdown not recovered
# # print("get_max_drawdown_underwater",underwater)
# # print("get_max_drawdown_underwater",underwater[:valley][underwater[:valley] == 0])
# # print("get_max_drawdown_underwater",peak, valley, recovery)
# # add a code,change index to datetime
# valley = list(underwater.index)[valley]
# return peak, valley, recovery
# 原版
valley = underwater.idxmin() # end of the period
# Find first 0
peak = underwater[:valley][underwater[:valley] == 0].index[-1]
# Find last 0
try:
recovery = underwater[valley:][underwater[valley:] == 0].index[0]
except IndexError:
recovery = np.nan # drawdown not recovered
# print(peak, valley, recovery)
return peak, valley, recovery
def get_max_drawdown(returns):
"""
Determines the maximum drawdown of a strategy.
Parameters
----------
returns : pd.Series
Daily returns of the strategy, noncumulative.
- See full explanation in :func:`~pyfolio.timeseries.cum_returns`.
Returns
-------
float
Maximum drawdown.
Note
-----
See https://en.wikipedia.org/wiki/Drawdown_(economics) for more details.
"""
returns = returns.copy()
df_cum = cum_returns(returns, 1.0)
running_max = np.maximum.accumulate(df_cum)
underwater = df_cum / running_max - 1
return get_max_drawdown_underwater(underwater)
def get_top_drawdowns(returns, top=10):
"""
Finds top drawdowns, sorted by drawdown amount.
Parameters
----------
returns : pd.Series
Daily returns of the strategy, noncumulative.
- See full explanation in tears.create_full_tear_sheet.
top : int, optional
The amount of top drawdowns to find (default 10).
Returns
-------
drawdowns : list
List of drawdown peaks, valleys, and recoveries. See get_max_drawdown.
"""
returns = returns.copy()
df_cum = ep.cum_returns(returns, 1.0)
running_max = np.maximum.accumulate(df_cum)
underwater = df_cum / running_max - 1
# print("df_cum",df_cum)
# print("running_max",running_max)
# print("underwater",underwater)
drawdowns = []
for t in range(top):
#print("len(underwater)",len(underwater))
peak, valley, recovery = get_max_drawdown_underwater(underwater)
# Slice out draw-down period
if not pd.isnull(recovery):
underwater.drop(underwater[peak: recovery].index[1:-1],
inplace=True)
else:
# drawdown has not ended yet
underwater = underwater.loc[:peak]
# print("get_top_drawdowns",peak, valley, recovery)
drawdowns.append((peak, valley, recovery))
if (len(returns) == 0) or (len(underwater) == 0):
break
# print(drawdowns)
return drawdowns
def gen_drawdown_table(returns, top=10):
"""
Places top drawdowns in a table.
Parameters
----------
returns : pd.Series
Daily returns of the strategy, noncumulative.
- See full explanation in tears.create_full_tear_sheet.
top : int, optional
The amount of top drawdowns to find (default 10).
Returns
-------
df_drawdowns : pd.DataFrame
Information about top drawdowns.
"""
df_cum = ep.cum_returns(returns, 1.0)
drawdown_periods = get_top_drawdowns(returns, top=top)
df_drawdowns = pd.DataFrame(index=list(range(top)),
columns=['Net drawdown in %',
'Peak date',
'Valley date',
'Recovery date',
'Duration'])
# print(df_drawdowns)
# print(drawdown_periods)
for i, (peak, valley, recovery) in enumerate(drawdown_periods):
if pd.isnull(recovery):
df_drawdowns.loc[i, 'Duration'] = np.nan
else:
df_drawdowns.loc[i, 'Duration'] = len(pd.date_range(peak,
recovery,
freq='B'))
# to_pydatetime()疑似是老的API，使用pd.to_datetime替代
# df_drawdowns.loc[i, 'Peak date'] = (peak.to_pydatetime()
# .strftime('%Y-%m-%d'))
# df_drawdowns.loc[i, 'Valley date'] = (valley.to_pydatetime()
# .strftime('%Y-%m-%d'))
# if isinstance(recovery, float):
# df_drawdowns.loc[i, 'Recovery date'] = recovery
# else:
# df_drawdowns.loc[i, 'Recovery date'] = (recovery.to_pydatetime()
# .strftime('%Y-%m-%d'))
df_drawdowns.loc[i, 'Peak date'] = (pd.to_datetime(peak).strftime('%Y-%m-%d'))
df_drawdowns.loc[i, 'Valley date'] = (pd.to_datetime(valley).strftime('%Y-%m-%d'))
if isinstance(recovery, float):
df_drawdowns.loc[i, 'Recovery date'] = recovery
else:
df_drawdowns.loc[i, 'Recovery date'] = (pd.to_datetime(recovery).strftime('%Y-%m-%d'))
df_drawdowns.loc[i, 'Net drawdown in %'] = (
(df_cum.loc[peak] - df_cum.loc[valley]) / df_cum.loc[peak]) * 100
df_drawdowns['Peak date'] = pd.to_datetime(df_drawdowns['Peak date'])
df_drawdowns['Valley date'] = pd.to_datetime(df_drawdowns['Valley date'])
df_drawdowns['Recovery date'] = pd.to_datetime(df_drawdowns['Recovery date'])
# print(df_drawdowns)
return df_drawdowns
def rolling_volatility(returns, rolling_vol_window):
"""
Determines the rolling volatility of a strategy.
Parameters
----------
returns : pd.Series
Daily returns of the strategy, noncumulative.
- See full explanation in tears.create_full_tear_sheet.
rolling_vol_window : int
Length of rolling window, in days, over which to compute.
Returns
-------
pd.Series
Rolling volatility.
"""
return returns.rolling(rolling_vol_window).std() \
* np.sqrt(APPROX_BDAYS_PER_YEAR)
def rolling_sharpe(returns, rolling_sharpe_window):
"""
Determines the rolling Sharpe ratio of a strategy.
Parameters
----------
returns : pd.Series
Daily returns of the strategy, noncumulative.
- See full explanation in tears.create_full_tear_sheet.
rolling_sharpe_window : int
Length of rolling window, in days, over which to compute.
Returns
-------
pd.Series
Rolling Sharpe ratio.
Note
-----
See https://en.wikipedia.org/wiki/Sharpe_ratio for more details.
"""
return returns.rolling(rolling_sharpe_window).mean() \
/ returns.rolling(rolling_sharpe_window).std() \
* np.sqrt(APPROX_BDAYS_PER_YEAR)
def simulate_paths(is_returns, num_days,
starting_value=1, num_samples=1000, random_seed=None):
"""
Gnerate alternate paths using available values from in-sample returns.
Parameters
----------
is_returns : pandas.core.frame.DataFrame
Non-cumulative in-sample returns.
num_days : int
Number of days to project the probability cone forward.
starting_value : int or float
Starting value of the out of sample period.
num_samples : int
Number of samples to draw from the in-sample daily returns.
Each sample will be an array with length num_days.
A higher number of samples will generate a more accurate
bootstrap cone.
random_seed : int
Seed for the pseudorandom number generator used by the pandas
sample method.
Returns
-------
samples : numpy.ndarray
"""
samples = np.empty((num_samples, num_days))
seed = np.random.RandomState(seed=random_seed)
for i in range(num_samples):
samples[i, :] = is_returns.sample(num_days, replace=True,
random_state=seed)
return samples
def summarize_paths(samples, cone_std=(1., 1.5, 2.), starting_value=1.):
"""
Gnerate the upper and lower bounds of an n standard deviation
cone of forecasted cumulative returns.
Parameters
----------
samples : numpy.ndarray
Alternative paths, or series of possible outcomes.
cone_std : list of int/float
Number of standard devations to use in the boundaries of
the cone. If multiple values are passed, cone bounds will
be generated for each value.
Returns
-------
samples : pandas.core.frame.DataFrame
"""
cum_samples = ep.cum_returns(samples.T,
starting_value=starting_value).T
cum_mean = cum_samples.mean(axis=0)
cum_std = cum_samples.std(axis=0)
if isinstance(cone_std, (float, int)):
cone_std = [cone_std]
cone_bounds = pd.DataFrame(columns=pd.Float64Index([]))
for num_std in cone_std:
cone_bounds.loc[:, float(num_std)] = cum_mean + cum_std * num_std
cone_bounds.loc[:, float(-num_std)] = cum_mean - cum_std * num_std
return cone_bounds
def forecast_cone_bootstrap(is_returns, num_days, cone_std=(1., 1.5, 2.),
starting_value=1, num_samples=1000,
random_seed=None):
"""
Determines the upper and lower bounds of an n standard deviation
cone of forecasted cumulative returns. Future cumulative mean and
standard devation are computed by repeatedly sampling from the
in-sample daily returns (i.e. bootstrap). This cone is non-parametric,
meaning it does not assume that returns are normally distributed.
Parameters
----------
is_returns : pd.Series
In-sample daily returns of the strategy, noncumulative.
- See full explanation in tears.create_full_tear_sheet.
num_days : int
Number of days to project the probability cone forward.
cone_std : int, float, or list of int/float
Number of standard devations to use in the boundaries of
the cone. If multiple values are passed, cone bounds will
be generated for each value.
starting_value : int or float
Starting value of the out of sample period.
num_samples : int
Number of samples to draw from the in-sample daily returns.
Each sample will be an array with length num_days.
A higher number of samples will generate a more accurate
bootstrap cone.
random_seed : int
Seed for the pseudorandom number generator used by the pandas
sample method.
Returns
-------
pd.DataFrame
Contains upper and lower cone boundaries. Column names are
strings corresponding to the number of standard devations
above (positive) or below (negative) the projected mean
cumulative returns.
"""
samples = simulate_paths(
is_returns=is_returns,
num_days=num_days,
starting_value=starting_value,
num_samples=num_samples,
random_seed=random_seed
)
cone_bounds = summarize_paths(
samples=samples,
cone_std=cone_std,
starting_value=starting_value
)
return cone_bounds
def extract_interesting_date_ranges(returns):
"""
Extracts returns based on interesting events. See
gen_date_range_interesting.
Parameters
----------
returns : pd.Series
Daily returns of the strategy, noncumulative.
- See full explanation in tears.create_full_tear_sheet.
Returns
-------
ranges : OrderedDict
Date ranges, with returns, of all valid events.
"""
returns_dupe = returns.copy()
returns_dupe.index = returns_dupe.index.map(pd.Timestamp)
ranges = OrderedDict()
for name, (start, end) in PERIODS.items():
try:
period = returns_dupe.loc[start:end]
if len(period) == 0:
continue
ranges[name] = period
except BaseException:
continue
return ranges
此处可能存在不合适展示的内容，页面不予展示。您可通过相关编辑功能自查并修改。
如您确认内容无涉及 不当用语 / 纯广告导流 / 暴力 / 低俗色情 / 侵权 / 盗版 / 虚假 / 无价值内容或违法国家有关法律法规的内容，可点击提交进行申诉，我们将尽快为您处理。
马建仓 AI 助手