# -*- coding: gbk -*-

##

# @file stock_plot.py

# @brief 一些绘制统计信息的函数。

# @author wondreamer

# @version 2.0

# @date 2015-10-19

import matplotlib

matplotlib.use("TKAgg")

from matplotlib.widgets import Cursor

from matplotlib.widgets import MultiCursor

from matplotlib.font_manager import FontProperties

import matplotlib.pyplot as plt

import numpy as np

import operator

import pandas as pd

#from pylab import *

#mpl.rcParams['font.sans-serif'] = ['Microsoft YaHei'] #指定默认字体

def accumulate(iterable, func=operator.add):

'Return running totals'

# accumulate([1,2,3,4,5]) --> 1 3 6 10 15

# accumulate([1,2,3,4,5], operator.mul) --> 1 2 6 24 120

it = iter(iterable)

total = next(it)

# rst = []

yield total

for element in it:

total = func(total, element)

yield total

font = FontProperties(size=8)

font_big = FontProperties(size=14)

class EventHandler(object):

"""docstring for EventHandler"""

def __init__(self, data=None, fig=None):

self.fig = fig

self.data = data

self.pre_x = None

def on_pick(self, event):

'''docstring for on_motion'''

print "888888"

print str(event.mouseevent.xdata)

#print event.artist

def on_move(self, event):

'''docstring for on_motion'''

if isinstance(event.xdata, np.float64):

i = int(event.xdata)/1

if self.pre_x != i:

print self.data.index[i]

print self.data[i]

c = pd.to_datetime(self.data.index[i])\

.strftime("%Y-%m-%d %H:%M:%S") + "\n" + "hh"

self.fig.axes[2].set_xlabel(c)

self.pre_x = i

def plot_simple_entry(fig, entry_nbar_best, entry_nbar_worst, nbar):

'''docstring for plot_simple_entry'''

fig.canvas.set_window_title(u'入场信息')

ax1 = fig.add_subplot(1, 1, 1)

entry_nbar_best = entry_nbar_best.order()

entry_nbar_worst = entry_nbar_worst.reindex(entry_nbar_best.index)

if len(entry_nbar_best) > 0:

entry_nbar_best.plot(

ax=ax1, kind='bar', color='red',

grid=False, use_index=False, label=u"%s根最优" % nbar)

#ax1.bar(range(len(entry_nbar_best)), entry_nbar_best, color='r', label=u"%s根最优"%nbar)

entry_nbar_worst.plot(

ax=ax1, kind='bar', color='y',

grid=False, use_index=False, label=u"%s根最差" % nbar)

temp = entry_nbar_worst[entry_nbar_worst < 0]

ax1.plot(range(len(entry_nbar_best)),

[temp.mean()]*len(entry_nbar_best),

'y--',

label=u"平均风险: %s" % temp.mean())

temp = entry_nbar_best[entry_nbar_best > 0]

ax1.plot(range(len(entry_nbar_best)),

[temp.mean()]*len(entry_nbar_best),

'r--',

label=u'平均最优: %s' % temp.mean())

ax1.legend(loc='upper left', prop=font).get_frame().set_alpha(0.5)

ax1.set_xticklabels([])

ax1.set_xlabel("")

return [ax1], []

def plot_entry(fig, exit_profit, entry_best, entry_worst,

entry_nbar_best, entry_nbar_worst, nbar, binwidth=1):

fig.canvas.set_window_title(u'入场信息')

axescolor = '#f6f6f6' # the axes background color

left, width = 0.1, 0.8

rect1 = [left, 0.7, width, 0.2] # left, bottom, width, height

rect2 = [left, 0.3, width, 0.4]

rect3 = [left, 0.1, width, 0.2]

ax1 = fig.add_axes(rect1, axisbg=axescolor)

ax2 = fig.add_axes(rect2, axisbg=axescolor, sharex=ax1)

ax3 = fig.add_axes(rect3, axisbg=axescolor, sharex=ax1)

(entry_best-exit_profit).plot(ax=ax1, kind='bar', grid=False,

use_index=False, label=u"优实差")

entry_worst.plot(ax=ax1, kind='bar', grid=False,

use_index=False, color='y', label=u"最大不利偏移")

if nbar > 0:

entry_nbar_best.plot(ax=ax3, kind='bar', color='red',

grid=False, use_index=False,

label=u"%s根最优" % nbar)

#ax3.bar(range(len(entry_nbar_best)), entry_nbar_best, color='r', label=u"%s根最优"%nbar)

entry_nbar_worst.plot(ax=ax3, kind='bar', color='y',

grid=False, use_index=False,

label=u"%s根最差" % nbar)

temp = entry_nbar_worst[entry_nbar_worst < 0]

ax3.plot(range(len(entry_nbar_best)),

[temp.mean()]*len(entry_nbar_best),

'y--',

label=u"平均风险: %s" % temp.mean())

temp = entry_nbar_best[entry_nbar_best > 0]

ax3.plot(range(len(entry_nbar_best)),

[temp.mean()]*len(entry_nbar_best),

'r--',

label=u'平均最优: %s' % temp.mean())

ax3.legend(loc='upper left', prop=font).get_frame().set_alpha(0.5)

for i in xrange(len(exit_profit)):

if entry_best[i] > 0 and exit_profit[i] > 0:

px21 = ax2.bar(i, exit_profit[i], width=binwidth, color='blue')

px22 = ax2.bar(i, entry_best[i]-exit_profit[i],

width=binwidth,

color='red',

bottom=exit_profit[i])

elif entry_best[i] < 0 and exit_profit[i] < 0:

ax2.bar(i, entry_best[i], width=binwidth, color='red')

ax2.bar(i, exit_profit[i]-entry_best[i],

width=binwidth, color='blue', bottom=entry_best[i])

else:

ax2.bar(i, entry_best[i], width=binwidth, color='red')

ax2.bar(i, exit_profit[i], width=binwidth, color='blue')

ax2.legend((px21[0], px22[0]), (u'实际盈利', u'早出最优盈利'),

loc='upper left',

prop=font).get_frame().set_alpha(0.5)

ax1.legend(loc='upper left', prop=font).get_frame().set_alpha(0.5)

ax1.set_ylabel(u"交易区间内的差值", fontproperties=font)

ax2.set_ylabel(u"交易区间内的盈利", fontproperties=font)

for ax in ax1, ax2, ax3:

# if ax!=ax3:

ax.set_xticklabels([])

ax3.set_xlabel("")

ax1.set_title(u"入场相关信息", fontproperties=font_big)

c1 = Cursor(ax2, useblit=True, color='red',

linewidth=1, vertOn=True, horizOn=True)

multi = MultiCursor(fig.canvas, fig.axes, color='r',

lw=1, horizOn=False, vertOn=True)

#handle = EventHandler(exit_profit, fig)

#fig.canvas.mpl_connect('motion_notify_event', handle.on_move)

#fig.canvas.mpl_connect('pick_event', handle.on_pick)

def format_coord(x, y):

""" 状态栏信息显示 """

i = int(x)/1

c = pd.to_datetime(exit_profit.index[i])\

.strftime("%Y-%m-%d %H:%M:%S") + \

" Profit: %s MAE: %s" % (exit_profit[i], entry_worst[i])

return str(c)

ax1.format_coord = format_coord

ax2.format_coord = format_coord

ax3.format_coord = format_coord

return [ax1, ax2, ax3], [multi, c1]

def plot_exit(fig, exit_profit, exit_nbar_best, exit_nbar_worst,

profits_more, risks, nbar, binwidth=1):

# fig.canvas.set_window_title(u'出场信息')

axescolor = '#f6f6f6' # the axes background color

left, width = 0.1, 0.8

rect2 = [left, 0.4, width, 0.4]

rect3 = [left, 0.1, width, 0.3]

ax1 = fig.add_axes(rect3, axisbg=axescolor)

ax2 = fig.add_axes(rect2, axisbg=axescolor, sharex=ax1)

if nbar > 0:

print "**66666"

# plot ax1

profits_more.plot(ax=ax1, kind='bar', grid = False, use_index = False, label=u"%s根最优"%nbar)

risks.plot(ax=ax1, kind='bar', grid = False, use_index = False, color = 'y', label=u"%s根最差"%nbar)

temp = risks[risks<0]

ax1.plot(range(len(temp)), [temp.mean()]*len(temp), 'y--', label=u"平均风险: %s"%temp.mean())

temp = profits_more[profits_more>0]

ax1.plot(range(len(temp)), [temp.mean()]*len(temp), 'r--', label=u"平均更优: %s"%temp.mean())

ax1.legend(prop=font, loc='upper left').get_frame().set_alpha(0.5)

#ax1.annotate(str(np.mean(risks)), xy=(len(records)/2, np.mean(risks)), xycoords='data',

#xytext=(-30, -30), textcoords='offset points', color='b',

#arrowprops=dict(arrowstyle="->",

#connectionstyle="arc3,rad=.2")

#)

# plot ax2

for i in xrange(len(exit_profit)):

if(exit_nbar_best[i]>exit_profit[i] and exit_profit[i]>0):

px21 = ax2.bar(i, exit_profit[i], width=binwidth, color='blue')

px22 = ax2.bar(i, exit_nbar_best[i]-exit_profit[i], width=binwidth, color='red', bottom = exit_profit[i])

elif(exit_nbar_best[i]<exit_profit[i] and exit_profit[i]>0 and exit_nbar_best[i]>0):

ax2.bar(i, exit_nbar_best[i], width=binwidth, color='red')

ax2.bar(i, exit_profit[i]-exit_nbar_best[i], width=binwidth, color='blue', bottom = exit_nbar_best[i])

elif(exit_nbar_best[i]<exit_profit[i] and exit_profit[i]<0):

ax2.bar(i, exit_profit[i], width=binwidth, color='red')

ax2.bar(i, exit_nbar_best[i]-exit_profit[i], width=binwidth, color='blue', bottom = exit_profit[i])

elif(exit_nbar_best[i]>exit_profit[i] and exit_profit[i]<0 and exit_nbar_best[i]<0):

ax2.bar(i, exit_nbar_best[i], width=binwidth, color='red')

ax2.bar(i, exit_profit[i]-exit_nbar_best[i], width=binwidth, color='blue', bottom = exit_nbar_best[i])

else:

ax2.bar(i, exit_nbar_best[i], width=binwidth, color='red')

ax2.bar(i, exit_profit[i], width=binwidth, color='blue')

ax2.legend((px21[0], px22[0]), (u'实际盈利', u'延出最优盈利'),loc='upper left', prop=font).get_frame().set_alpha(0.5)

ax2.set_ylabel(u"交易区间内的盈利", fontproperties = font)

for ax in ax1, ax2:

#if ax!=ax1:

ax.set_xticklabels([])

ax1.set_xlabel("")

ax2.set_title(u"出场相关信息", fontproperties=font_big)

multi = MultiCursor(fig.canvas, fig.axes, color='r', lw=1, horizOn=False, vertOn=True)

return [ax1, ax2], [multi]

else:

return [], []

def plot_summary(fig, exit_profit, entry_best, entry_worst, entry_nbar_best, entry_nbar_worst,

exit_nbar_best, exit_nbar_worst, profits_more, risks, NBAR):

fig.canvas.set_window_title(u'画图汇总')

ax11 = fig.add_subplot(3, 2, 1)

ax12 = fig.add_subplot(3, 2, 2)

ax21 = fig.add_subplot(3, 2, 3)

ax22 = fig.add_subplot(3, 2, 4)

ax31 = fig.add_subplot(3, 2, 5)

ax32 = fig.add_subplot(3, 2, 6)

#plt.subplots_adjust(left=0, right=1)

# Profits Distribution

shift = pd.Series([0]*len(exit_profit[exit_profit<=0]))

temp = pd.concat([shift, exit_profit[exit_profit>0]])

temp.index = range(len(temp))

temp.plot(ax=ax11, grid=False, use_index=False, style="r", label=u'盈利')

ax11.fill_between(range(len(temp)), [0]*len(temp), temp.tolist(), facecolor='r')

temp = 0 - exit_profit[exit_profit<=0]

ax11.plot(temp, 'y', label=u'亏损')

ax11.fill_between(range(len(temp)), [0]*len(temp), temp.tolist(), facecolor='y')

ax11.plot(entry_worst, 'b', label=u'最差偏移')

ax11.axhline(color='black')

ax11.legend(prop=font, loc='upper left').get_frame().set_alpha(0.5)

# Profits Distribution Bins

#exit_profit.hist(ax=ax12, bins=50, normed=True, color='r')

#n, bins = np.histogram(exit_profit.tolist(), 50, normed=True)

#ax12.plot([0, 0], [0, max(n)], color='y', linewidth=2)

#ax12.grid(False)

exit_profit.plot(ax=ax12, kind='kde', color='b', label="")

binwidth = abs(exit_profit.min()/9)

bins = np.arange(exit_profit.min(), exit_profit.max() + binwidth, binwidth)

ax12.hist(exit_profit[exit_profit>0], bins=bins, color = 'red' , normed=False, label=u'盈利分布')

ax12.hist(exit_profit[exit_profit<0], bins=bins, color = 'y' , normed=False, label=u'亏损分布')

plot_contribution(ax12, bins, exit_profit, 'bo--')

ax12.legend(prop=font, loc='upper left').get_frame().set_alpha(0.5)

#ax12.set_yscale('log')

# MAE

MAE = entry_worst.reindex(exit_profit[exit_profit>0].index)

MAE.order().plot(ax=ax21,style='r', grid=False, use_index=False, label=u'最大不利偏移')

exit_profit[exit_profit<0].plot(ax=ax21, style='y', grid=False, use_index=False, label=u'亏损分布')

worst = MAE.min()

print u"最大不利偏移: %s" % worst

bb = exit_profit[exit_profit<0]

aa = [worst]*len(bb)

ax21.fill_between(range(len(exit_profit[exit_profit<0])), aa, bb, where=bb<aa, color='red')

ax21.set_ylim((min(exit_profit.min(), MAE.min())-10), 0)

ax21.legend(prop=font, loc='upper left').get_frame().set_alpha(0.5)

# Potential Profits When Lose

temp = entry_best.reindex(exit_profit[exit_profit<0].index)

ax22.plot(temp.tolist(), color='r', label=u"最优盈利" )

ax22.fill_between(range(len(temp)), temp.tolist(), [0]*len(temp), facecolor='r')

ax22.plot(temp.order().tolist(), color='b', label=u"有序最优盈利" )

ax22.plot(exit_profit[exit_profit<0].tolist(), color='y', label=u'实际亏损')

ax22.set_ylim((min(exit_profit.min(), MAE.min())-10, temp.max()+10))

ax22.legend(prop=font, loc='upper left').get_frame().set_alpha(0.5)

ax22.axhline(0, c='black')

if NBAR > 0:

# Entry N Bar

enbest = entry_nbar_best.reindex(entry_nbar_best[entry_nbar_best>0].index).order()

enbest.plot(ax=ax31,style='r', grid=False, use_index=False, label=u"%s根最优平均: %s"%(NBAR, entry_nbar_best[entry_nbar_best>0].mean()))

enworst = (0-entry_nbar_worst.reindex(entry_nbar_worst[entry_nbar_worst<0].index).order(ascending=False))

enworst.plot(ax=ax31, style='y', grid=False, use_index=False,

label=u"%s根最差平均: %s"%(NBAR, entry_nbar_worst[entry_nbar_worst<0].mean()))

ax31.axhline(0, c='black')

ax31.legend(prop=font, loc='upper left').get_frame().set_alpha(0.5)

# Exit N Bar

profits_more.reindex(profits_more[profits_more>0].index).order().plot(ax=ax32,style='r',

grid=False, use_index=False, label=u"%s根最优平均: %s"%(NBAR, profits_more[profits_more>0].mean()))

(0-risks.reindex(risks[risks<0].index).order(ascending=False)).plot(ax=ax32, style='y',

grid=False, use_index=False, label=u"%s根最差平均: %s"%(NBAR,risks[risks<0].mean()))

ax32.legend(prop=font, loc='upper left').get_frame().set_alpha(0.5)

#

#ax31.xaxis_date()

map(lambda x: x.set_xticklabels([]), [ax11, ax21, ax22, ax31, ax32])

map(lambda x: x.set_xlabel(""), [ax11, ax12, ax21, ax22, ax31, ax32])

map(lambda x: x.set_ylabel(""), [ax11, ax12, ax21, ax22, ax31, ax32])

ax11.set_xlabel(u"盈利分布", fontproperties=font_big)

ax12.set_ylabel(u"盈利统计", fontproperties=font_big)

ax21.set_xlabel(u"最大不利偏移和亏损", fontproperties=font_big)

ax22.set_xlabel(u"亏损交易的潜在盈利空间", fontproperties=font_big)

ax31.set_xlabel(u"进场后%s根"%NBAR, fontproperties=font_big)

ax32.set_xlabel(u"离场后%s根"%NBAR, fontproperties=font_big)

cursors = []

for ax in [ax11, ax12, ax21, ax22, ax31, ax32]:

cursors.append(Cursor(ax, useblit=True, color='red', linewidth=1,

vertOn = True, horizOn = True))

return [ax11, ax12, ax21, ax22, ax31, ax32], cursors

def plot_scatter(fig, x, y, x2, y2, binnum):

'''docstring for plot_test'''

fig.canvas.set_window_title(u'交易鸟瞰图')

# definitions for the axes

left, width = 0.1, 0.65

bottom, height = 0.1, 0.65

bottom_h = left_h = left+width+0.02

rect_scatter = [left, bottom, width, height]

rect_histx = [left, bottom_h, width, 0.2]

rect_histy = [left_h, bottom, 0.2, height]

# start with a rectangular Figure

axScatter = plt.axes(rect_scatter)

axHistx = plt.axes(rect_histx)

axHisty = plt.axes(rect_histy)

cursor = Cursor(axScatter, useblit=True, color='red', linewidth=1 )

axScatter.plot(x, y, 'o', color = 'red')

axScatter.plot(x2, y2, 'o', color = 'blue')

# now determine nice limits by hand:

xmax = np.max(x+x2)

xmin = np.min(x+x2)

binwidth = xmax / binnum

lim = ( int(xmax/binwidth) + 1) * binwidth

bins = np.arange(-lim, lim + binwidth, binwidth)

axHistx.hist(x+x2, bins=bins)

ymax = np.max(y+y2)

ymin = np.min(y+y2)

binwidth = ymax/binnum

lim = ( int(ymax/binwidth) + 1) * binwidth

bins = np.arange(-lim, lim + binwidth, binwidth)

axHisty.hist(y, bins=bins, orientation='horizontal', color = 'red' )

axHisty.hist(y2, bins=bins, orientation='horizontal', color = 'blue' )

xymax = np.max( [np.max(np.fabs(x+x2)), np.max(np.fabs(y+y2))] )

lim = ( int(xymax/binwidth) + 1) * binwidth

axScatter.axhline(color='black')

#axScatter.set_xlim( (-xmin-10, xmax+10))

#axScatter.set_ylim((-ymin-10, ymax+10))

axHistx.set_xlim( axScatter.get_xlim() )

axHisty.set_ylim( axScatter.get_ylim() )

axHisty.set_xlabel(u"盈亏分布", fontproperties = font_big)

axHistx.set_ylabel(u"周期分布", fontproperties = font_big)

axScatter.set_xlabel(u"盈亏和周期分布", fontproperties = font_big)

axScatter.grid(True)

axHistx.grid(True)

axHisty.grid(True)

c = Cursor(axScatter, useblit=True, color='red', linewidth=1, vertOn = True, horizOn = True)

return [axScatter, axHistx, axHisty], [c]

def plot_compare(exit_profits, entry_bests, entry_worsts, entry_nbar_bests, entry_nbar_worsts,

exit_nbar_bests, exit_nbar_worsts, profits_mores, risks, colors, names, NBAR):

fig = plt.figure(facecolor='white')

fig.canvas.set_window_title(u'画图汇总一')

ax11 = fig.add_subplot(3, 2, 1)

ax12 = fig.add_subplot(3, 2, 2)

ax21 = fig.add_subplot(3, 2, 3)

ax22 = fig.add_subplot(3, 2, 4)

ax31 = fig.add_subplot(3, 2, 5)

ax32 = fig.add_subplot(3, 2, 6)

#plt.subplots_adjust(left=0, right=1)

for i in range(len(exit_profits)):

nm = names[i]

exit_profit = exit_profits[i]

entry_best = entry_bests[i]

entry_worst = entry_worsts[i]

entry_nbar_best = entry_nbar_bests[i]

entry_nbar_worst = entry_nbar_worsts[i]

exit_nbar_best = exit_nbar_bests[i]

exit_nbar_worst = exit_nbar_worsts[i]

profits_more = profits_mores[i]

risk = risks[i]

c = colors[i]

# Profits Distribution

shift = pd.Series([0]*len(exit_profit[exit_profit<=0]))

temp = pd.concat([shift, exit_profit[exit_profit>0]])

temp.index = range(len(temp))

temp.plot(ax=ax11, grid=False, use_index=False, style=c, label=u'%s盈利'%nm)

temp = 0 - exit_profit[exit_profit<=0]

ax11.plot(temp.tolist(), c, label=u'%s亏损'%nm)

ax11.plot(entry_worst.tolist(), c, label=u'%s最差偏移'%nm)

#ax11.set_xscale('log')

# Profits Distribution Bins

#exit_profit.hist(ax=ax12, bins=50, normed=True, color=c)

#exit_profit.plot(ax=ax12, kind='kde', color='b', label="")

#ax12.legend(prop=font, loc='upper left').get_frame().set_alpha(0.5)

a = np.histogram(exit_profit.tolist(), 50, normed=True)

n = pd.Series(a[0])

bins = pd.Series(a[1][:-1])

temp = bins[bins>0]

ax12.plot(temp.tolist(), n.reindex(temp.index).tolist(), c, label=u'%s盈利分布'%nm)

temp = bins[bins<0]

ax12.plot(temp.tolist(), n.reindex(temp.index).tolist(), '%s--'%c, label=u'%s亏损分布'%nm)

ax12.legend(prop=font, loc='upper left').get_frame().set_alpha(0.5)

# MAE

MAE = entry_worst.reindex(exit_profit[exit_profit>0].index)

MAE.order().plot(ax=ax21,style=c, grid=False, use_index=False, label=u'%s最大不利偏移'%nm)

exit_profit[exit_profit<0].plot(ax=ax21, style='%s--'%c, grid=False, use_index=False, label=u'%s亏损分布'%nm)

# Potential Profits When Lose

temp = entry_best.reindex(exit_profit[exit_profit<0].index)

ax22.plot(temp.tolist(), c, label=u"%s最优盈利" % nm)

ax22.plot(temp.order().tolist(), '%s--'%c, label=u"%s有序最优盈利" % nm)

ax22.plot(exit_profit[exit_profit<0].tolist(), '%s--'%c, label=u'%s实际亏损'%nm)

if len(entry_nbar_best)>0:

# Entry N Bar

entry_nbar_best.reindex(entry_nbar_best[entry_nbar_best>0].index).order().plot(ax=ax31,style=c,

grid=False, use_index=False, label=u"%s%s根最优平均: %s"%(nm, NBAR, entry_nbar_best[entry_nbar_best>0].mean()))

(0-entry_nbar_worst.reindex(entry_nbar_worst[entry_nbar_worst<0].index).order(ascending=False)).plot(ax=ax31, style='%s--'%c,

grid=False, use_index=False, label=u"%s%s根最差平均: %s"%(nm, NBAR, entry_nbar_worst[entry_nbar_worst<0].mean()))

# Exit N Bar

profits_more.reindex(profits_more[profits_more>0].index).order().plot(ax=ax32,style=c,

grid=False, use_index=False, label=u"%s%s根最优平均: %s"%(nm,NBAR, profits_more[profits_more>0].mean()))

(0-risk.reindex(risk[risk<0].index).order(ascending=False)).plot(ax=ax32, style='%s--'%c,

grid=False, use_index=False, label=u"%s%s根最差平均: %s"%(nm,NBAR,risk[risk<0].mean()))

#

#ax31.xaxis_date()

map(lambda x: x.set_xticklabels([]), [ax11, ax21, ax22, ax31, ax32])

map(lambda x: x.set_xlabel(""), [ax11, ax12, ax21, ax22, ax31, ax32])

map(lambda x: x.set_ylabel(""), [ax11, ax12, ax21, ax22, ax31, ax32])

ax11.set_xlabel(u"盈利分布", fontproperties=font_big)

ax12.set_ylabel(u"盈利统计", fontproperties=font_big)

ax12.axvline(color='black')

ax21.legend(prop=font, loc='upper left').get_frame().set_alpha(0.5)

ax21.set_xlabel(u"最大不利偏移和亏损", fontproperties=font_big)

ax22.set_xlabel(u"亏损交易的潜在盈利空间", fontproperties=font_big)

ax31.set_xlabel(u"进场后%s根"%NBAR, fontproperties=font_big)

ax32.set_xlabel(u"离场后%s根"%NBAR, fontproperties=font_big)

ax11.axhline(color='black')

ax11.legend(prop=font, loc='upper left').get_frame().set_alpha(0.5)

ax22.legend(prop=font, loc='upper left').get_frame().set_alpha(0.5)

ax22.axhline(0, c='black')

ax31.axhline(0, c='black')

ax31.legend(prop=font, loc='upper left').get_frame().set_alpha(0.5)

ax32.legend(prop=font, loc='upper left').get_frame().set_alpha(0.5)

ax12.set_xlim((np.min(a[1][:-1])-100, np.max(a[1][:-1])+50))

ax21.set_ylim((min(exit_profit.min(), MAE.min())-10), 0)

ax22.set_ylim((min(exit_profit.min(), MAE.min())-10, temp.max()+10))

cursors = []

for ax in [ax11, ax12, ax21, ax22, ax31, ax32]:

cursors.append(Cursor(ax, useblit=True, color='red', linewidth=1,

vertOn = True, horizOn = True))

return fig, cursors

def ax_normed_data(x1list, y1list, ax_ymax):

'''docstring for normed_data'''

unit = ax_ymax / max(abs(y1list))

nxlist = []

nylist = []

for i in range(len(y1list)):

if y1list[i] != 0:

nxlist.append(x1list[i])

nylist.append(y1list[i])

nylist = np.abs(np.array(nylist)*unit)

return nxlist, nylist

def plot_contribution(ax, bins, v, style):

'''docstring for plot_contribution'''

ctri = np.array(range(len(bins)-1))

ymin, ymax = ax.get_ylim()

for i in range(len(bins)-1):

t = v[bins[i]<=v]

t = t[t<bins[i+1]]

ctri[i] = t.sum()

x = [(bins[i]+bins[i+1])/2 for i in range(len(bins)-1)]

nx, ny = ax_normed_data(x, ctri, ymax)

ax.plot(nx, ny, style)

losex = []

losey = []

winx = []

winy = []

for i in range(len(x)):

if x[i]<0:

losex.append(x[i])

losey.append(ctri[i])

else:

winx.append(x[i])

winy.append(ctri[i])

losey = [i for i in accumulate(losey[::-1])]

losey = losey[::-1]

winy = [i for i in accumulate(winy)]

nx, ny = ax_normed_data(x, np.array(losey+winy), ymax)

ax.plot(nx, ny, 'k')

def plot_summary2(fig, rtn, entry_best, data_win, data_lose, exit_profit,

exit_nbar_best, exit_nbar_worst, nbar):

''' loseed arguments '''

cursors = []

winrtn = rtn.reindex(data_win.index)

losertn = rtn.reindex(data_lose.index)

fig.canvas.set_window_title(u'画图汇总二')

ax11 = fig.add_subplot(2, 2, 1)

ax11.plot(range(len(losertn)), losertn.tolist(), 'yo--', label=u'亏损回撤')

ax11.plot(len(losertn)+np.arange(len(winrtn)), winrtn.tolist(), 'ro--', label=u'盈利回撤')

ax11.plot(rtn.order().tolist(), 'b')

ax11.legend(prop=font, loc='upper left').get_frame().set_alpha(0.5)

cursors.append(Cursor(ax11, useblit=True, color='red', linewidth=1,

vertOn = True, horizOn = True))

ax11.set_xlabel(u'回撤', fontproperties=font_big)

ax12 = fig.add_subplot(2, 2, 2)

binwidth = (rtn.max() - rtn.min()) / 30

#rtn.plot(ax=ax12, kind='kde', color='b', label="")

bins = np.arange(rtn.min(), rtn.max() + binwidth, binwidth)

rst = ax12.hist(rtn, bins=bins, color = 'y' , normed=False, label=u'回撤分布')

n, bins = rst[0], rst[1]

plot_contribution(ax12, bins, rtn, 'bo--')

ax12.legend(prop=font, loc='upper left').get_frame().set_alpha(0.5)

cursors.append(Cursor(ax12, useblit=True, color='red', linewidth=1,

vertOn = True, horizOn = True))

#ax21 = fig.add_subplot(3, 2, 3)

#ds = entry_best.reindex(data_lose.index)-data_lose['exit_profit']

#ax21.plot(range(len(ds)), ds, 'yo--', label=u'亏损回吐')

#dl = (entry_best.reindex(data_win.index)-data_win['exit_profit']).tolist()

#ax21.plot(len(ds)+np.arange(len(dl)), dl, 'ro--', label=u'盈利回吐')

#ax21.set_xticklabels([])

#ax21.set_xlabel(u'回吐', fontproperties=font_big)

#ax21.legend(prop=font, loc='upper left').get_frame().set_alpha(0.5)

#cursors.append(Cursor(ax21, useblit=True, color='red', linewidth=1,

#vertOn = True, horizOn = True))

#diff = entry_best - exit_profit

#ax22 = fig.add_subplot(3, 2, 4)

#binwidth = (diff.max() - diff.min()) / 30

##diff.plot(ax=ax22, kind='kde', color='b', label="")

#bins = np.arange(diff.min(), diff.max() + binwidth, binwidth)

#rst = ax22.hist(diff, bins=bins, color = 'y' , normed=False, label=u'回吐分布')

#n, bins = rst[0], rst[1]

#plot_contribution(ax22, bins, diff, 'bo--')

#ax22.legend(prop=font, loc='upper left').get_frame().set_alpha(0.5)

#cursors.append(Cursor(ax22, useblit=True, color='red', linewidth=1,

#vertOn = True, horizOn = True))

if nbar>0:

ax31 = fig.add_subplot(2, 2, 3)

bl = (exit_nbar_best.reindex(data_lose.index)-data_lose['exit_profit']).order(ascending=False)

wl = (exit_nbar_worst.reindex(data_lose.index)-data_lose['exit_profit']).reindex(bl.index)

ax31.plot(bl, color='y')

ax31.plot(wl, color='k')

ax31.fill_between(range(len(bl)), bl, wl, facecolor='y')

bw = (exit_nbar_best.reindex(data_win.index)-data_win['exit_profit']).order()

ww = (exit_nbar_worst.reindex(data_win.index)-data_win['exit_profit']).reindex(bw.index)

ax31.plot(len(bl)+np.arange(len(bw)), bw.tolist(), 'r')

ax31.plot(len(bl)+np.arange(len(ww)), ww.tolist(), 'k')

ax31.fill_between(len(bl)+np.arange(len(ww)), bw, ww, facecolor='r', label=u'hello')

ax31.plot(data_lose['exit_profit'].abs().reindex(bl.index), 'b')

cursors.append(Cursor(ax31, useblit=True, color='red', linewidth=1,

vertOn = True, horizOn = True))

ax31.axhline(color='k')

return [ax11, ax12, ax31], cursors