用python做股票量化分析豆瓣_[转]构建基于R的交易系统(5)quantstrat包（中）（来源：豆瓣-数据铺子）...-CSDN博客

生命不息，学习不止~

尊重原创，原帖地址(豆瓣-数据铺子)：

感谢数据铺子的“构建基于R的交易系统”系列帖子。

--- faruto按

(2)基于quantstrat包的回测：单资产情况

现在可以结合quantstrat包和blotter包给出一个更好的量化策略回测。首先做初始化，这里包括blotter包里边对金融工具和账户和组合的初始化，也包括quantstrat包对策略和指令的初始化。

下面以上一篇用过的faber策略为例。

library(quantstrat)

金融产品初始化

currency("RMB")

## [1]

"RMB"

stock("ZSYH", currency = "RMB", multiplier = 1)

## [1]

"ZSYH"

设定时区

Sys.setenv(TZ = "UTC")

读取金融交易数据并转化为月数据

ZSYH

getSymbols("600036.ss", from = "2008-01-01", to = Sys.Date(), src =

"yahoo",

auto.assign = FALSE)

## As of

0.4-0, 'getSymbols' uses env=parent.frame() and auto.assign=TRUE

default.

## This

behavior will be phased out in 0.5-0 when the call will default

## use

auto.assign=FALSE. getOption("getSymbols.env") and

getOptions("getSymbols.auto.assign") are now checked for

alternate

defaults

## This

message is shown once per session and may be disabled by

setting

options("getSymbols.warning4.0"=FALSE). See ?getSymbol for more

details

ZSYH

to.monthly(ZSYH, indexAt = "endof")

ZSYH$SMA10m

head(ZSYH$SMA10m)

SMA10m

2008-01-31 NA

2008-02-29 NA

2008-03-31 NA

2008-04-30 NA

2008-05-30 NA

2008-06-30 NA

初始化组合和账户

q.strategy

initPortf(q.strategy, "ZSYH", initDate =

"2007-12-31")

## [1]

"qFaber"

initAcct(q.strategy, portfolios = q.strategy, initDate =

"2007-12-31", initEq = 1e+06)

## [1]

"qFaber"

初始化指定和策略

initOrders(portfolio = q.strategy, initDate =

"2007-12-31")

strategy(q.strategy, store = TRUE)

ls(all =

T) #quantstrat创建了.strategy环境

## [1]

".blotter" ".strategy" "q.strategy" "ZSYH"

策略是什么呢？看一下

strategy

summary(strategy)

## Length

Class Mode

## name 1

-none- character

## assets

0 -none- NULL

indicators 0 -none- list

## signals

0 -none- list

## rules 1

-none- list

constraints 0 -none- NULL

## init 0

-none- list

## wrapup

0 -none- list

## call 3

-none- call

下面是quantstrat包的关键：加入指标、信号和规则。

加入一个指标，10月均线

add.indicator(strategy = q.strategy, name = "SMA",

arguments = list(x = quote(Cl(mktdata)),

n =

10), label = "SMA10")

## [1]

"qFaber"

加入信号，向上交叉10月线，向下交叉10月线

add.signal(q.strategy, name = "sigCrossover", arguments =

list(columns = c("Close",

"SMA10"), relationship = "gt"), label = "Cl.gt.SMA")

## [1]

"qFaber"

add.signal(q.strategy, name = "sigCrossover", arguments =

list(columns = c("Close",

"SMA10"), relationship = "lt"), label = "Cl.lt.SMA")

## [1]

"qFaber"

加入规则，买入规则和卖出规则

add.rule(q.strategy, name = "ruleSignal", arguments =

list(sigcol = "Cl.gt.SMA",

sigval = TRUE, orderqty = 900, ordertype = "market", orderside =

"long",

pricemethod = "market"), type = "enter", path.dep = TRUE) #

买入数量为900股

## [1]

"qFaber"

add.rule(q.strategy, name = "ruleSignal", arguments =

list(sigcol = "Cl.lt.SMA",

sigval = TRUE, orderqty = "all", ordertype = "market", orderside =

"long",

pricemethod = "market"), type = "exit", path.dep = TRUE)

## [1]

"qFaber"

此时的策略对象

summary(getStrategy(q.strategy))

## Length

Class Mode

## name 1

-none- character

## assets

0 -none- NULL

indicators 1 -none- list

## signals

2 -none- list

## rules 3

-none- list

constraints 0 -none- NULL

## init 0

-none- list

## wrapup

0 -none- list

## call 3

-none- call

现在的策略strategy中有1个指标,2种信号和相应的两个规则(买入、卖出)。

下面可以使用这个策略了。

out

applyStrategy(strategy = q.strategy, portfolios =

q.strategy)

## [1]

"2009-05-29 00:00:00 ZSYH 900 @ 12.97"

## [1]

"2010-01-29 00:00:00 ZSYH -900 @ 15.17"

## [1]

"2010-10-29 00:00:00 ZSYH 900 @ 14.57"

## [1]

"2010-11-30 00:00:00 ZSYH -900 @ 13.05"

## [1]

"2011-03-31 00:00:00 ZSYH 900 @ 14.09"

## [1]

"2011-06-30 00:00:00 ZSYH -900 @ 13.02"

## [1]

"2012-01-31 00:00:00 ZSYH 900 @ 12.65"

## [1]

"2012-03-30 00:00:00 ZSYH -900 @ 11.9"

## [1]

"2012-04-30 00:00:00 ZSYH 900 @ 12.2"

## [1]

"2012-05-31 00:00:00 ZSYH -900 @ 11.66"

## [1]

"2012-12-31 00:00:00 ZSYH 900 @ 13.75"

## [1]

"2013-06-28 00:00:00 ZSYH -900 @ 11.6"

summary(out)

## Length

Class Mode

## qFaber

1 -none- list

mktdata["2013"]

ZSYH.Open ZSYH.High ZSYH.Low ZSYH.Close ZSYH.Volume

2013-01-31 13.75 14.80 13.16 14.30 2.467e+09

2013-02-28 14.20 15.01 12.58 13.30 1.906e+09

2013-03-29 13.30 13.60 11.70 12.63 1.979e+09

2013-04-30 12.49 12.98 12.01 12.15 9.797e+08

2013-05-31 12.15 14.10 11.95 13.58 1.909e+09

2013-06-28 13.58 13.89 10.00 11.60 1.253e+09

2013-07-31 11.50 12.43 10.40 10.60 1.322e+09

2013-08-30 10.69 12.14 10.35 10.66 1.257e+09

2013-09-30 10.66 12.19 10.61 10.92 1.659e+09

2013-10-31 10.92 11.10 10.55 10.74 1.061e+09

2013-11-01 10.87 10.95 10.79 10.91 6.854e+07

ZSYH.Adjusted ZSYH.Close.SMA.10.SMA10 Cl.gt.SMA

Cl.lt.SMA

2013-01-31 13.68 11.31 NA NA

2013-02-28 12.73 11.42 NA NA

2013-03-29 12.09 11.52 NA NA

2013-04-30 11.63 11.64 NA NA

2013-05-31 13.00 12.00 NA NA

2013-06-28 11.60 12.16 NA 1

2013-07-31 10.60 12.20 NA NA

2013-08-30 10.66 12.26 NA NA

2013-09-30 10.92 12.35 NA NA

2013-10-31 10.74 12.05 NA NA

2013-11-01 10.91 11.71 NA NA

mktdata是在执行策略过程中创建的一个特殊的变量，包含了原交易数据的时间序列以及指标和信号。在我们上面的ZSYH里，在2013年6月出现一次卖出的信号。

接下来对相关对象(组合、账户和权益价值)进行更新。

updatePortf(q.strategy)

Warning: instrument USD not found, please create it first.

Warning:

CurrencyUSD not found, using currency multiplier of 1

## [1]

"qFaber"

updateAcct(q.strategy)

## [1]

"qFaber"

updateEndEq(q.strategy)

## [1]

"qFaber"

最后按照现在的组合和账户情况给出策略表现

myTheme

myTheme$col$dn.col

myTheme$col$up.col

myTheme$col$dn.border

myTheme$col$up.border

策略表现可视化

chart.Posn(q.strategy, Symbol = "ZSYH", Dates = "2008::",

theme = myTheme)

# 交易统计

(tstats

"ZSYH"))

Portfolio Symbol Num.Txns Num.Trades Net.Trading.PL

Avg.Trade.PL

## ZSYH

qFaber ZSYH 12 6 -3447 -574.5

Med.Trade.PL Largest.Winner Largest.Loser Gross.Profits

Gross.Losses

## ZSYH

-819 1980 -1935 1980 -5427

Std.Dev.Trade.PL Percent.Positive Percent.Negative

Profit.Factor

## ZSYH

1355 16.67 83.33 0.3648

Avg.Win.Trade Med.Win.Trade Avg.Losing.Trade

Med.Losing.Trade

## ZSYH

1980 1980 -1085 -963

Avg.Daily.PL Med.Daily.PL Std.Dev.Daily.PL Max.Drawdown

## ZSYH

-574.5 -819 1355 -9450

Profit.To.Max.Draw Avg.WinLoss.Ratio Med.WinLoss.Ratio

Max.Equity

## ZSYH

-0.3648 1.824 2.056 6003

Min.Equity End.Equity

## ZSYH

-3447 -3447

指令簿(order

book)

getOrderBook(q.strategy)

head(ob$qFaber$ZSYH)

Order.Qty Order.Price Order.Type Order.Side

Order.Threshold

2009-05-29 "900" "12.97" "market" "long" NA

2010-01-29 "all" "15.17" "market" "long" NA

2010-10-29 "900" "14.57" "market" "long" NA

2010-11-30 "all" "13.05" "market" "long" NA

2011-03-31 "900" "14.09" "market" "long" NA

2011-06-30 "all" "13.02" "market" "long" NA

Order.Status Order.StatusTime Prefer Order.Set Txn.Fees

2009-05-29 "closed" "2009-06-30 00:00:00" "" NA "0"

2010-01-29 "closed" "2010-02-26 00:00:00" "" NA "0"

2010-10-29 "closed" "2010-11-30 00:00:00" "" NA "0"

2010-11-30 "closed" "2010-12-31 00:00:00" "" NA "0"

2011-03-31 "closed" "2011-04-29 00:00:00" "" NA "0"

2011-06-30 "closed" "2011-07-29 00:00:00" "" NA "0"

Rule

2009-05-29 "ruleSignal.rule"

2010-01-29 "ruleSignal.rule"

2010-10-29 "ruleSignal.rule"

2010-11-30 "ruleSignal.rule"

2011-03-31 "ruleSignal.rule"

2011-06-30 "ruleSignal.rule"

quantstrat包给出了MAE(maximum adverse

excursion)和MFE(maximum favorable

excursion)的图形

chart.ME(Portfolio = q.strategy, Symbol = "ZSYH", type =

"MAE", scale = "percent")

chart.ME(Portfolio = q.strategy, Symbol = "ZSYH", type =

"MFE", scale = "percent")

和上一篇blooter里同一个策略的回测过程作比较，quantstrat包的特点是在很大程度上标准化了交易的策略实现，把策略分解为指标、信号和规则这三个部分。对于基于技术交易指标构造的量化策略来说，这种标准化当然是很方便的。特别是需要自己构造复杂的交易策略的时候。只要把注意力集中于策略本身就可以了。

当然，blotter包和quantstrat依然有很多使用上的不方便(我这里把运行时的各种警告都忽略了，呵呵)的地方。所以我想了解或者学习这些工具更重要的意义在于：他山之石，可以攻玉。借鉴别人的工作，丰富自己的工作。

下一篇继续写quantstrat，多资产的回测。