解决方案，请在函数装饰器上进行修，添加参数requuire_exists：
@preprocess(engine=coerce_string_to_eng(require_exists=False))

原因：
utils/sqlite.py修正API导致。

 76         raw_arrays = self.raw_price_loader.load_raw_arrays(

---> 77 colnames,
78 start_date,
79 end_date,

的结果是272 ×3040
---> 82 adjustments = self.adjustments_loader.load_adjustments(
83 colnames,
84 dates,
之后运行到 for c, c_raw, c_adjs in zip(columns, raw_arrays, adjustments):
时 c_adjs对应的矩阵是279 × 3040
不匹配，于是报错 ValueError: Mask shape (279, 3040) != data shape (272, 3040).

读取数据的时限为 2017-1-1 到2017-10-30 ， 用的bundle是tdx注入的

https://www.visualstudio.com/downloads/#build-tools-for-visual-studio-2017

在安装时，因为软件默认已经选择了需要安装的MSBuild组件，所以直接安装，不需要另选组件。

若依然失败，运行：

pip install --upgrade setuptools

实盘的讨论issues

欢迎大家踊跃发言，提供意见

1. 接口

目前实盘采用trade.dll接口，本项目不提供dll，请自行百度/谷歌，如需技术支持，请联系434588628

2. 具体实现

1. zipline-live：
   zipline-live 项目扩展了zipline，使用国外的ib broker接口来进行实盘对接，所以cn_zipline修改zipline-live，通过trade.dll来实现实盘对接。

2. dll的局限：
   由于trade.dll只有win32版本，并且dll本身只支持c/c++语言，由于我之前已经写了一个python混编的接口调用trade.dll，所以目前打算直接使用这个混编接口，这也就限定了python的环境必须是win32。

   由于cn_zipline的初衷是一个如zipline一样的跨平台回测以及实盘框架，而trade.dll的win32版本要求python只能是win32版本。为脱离win32的束缚，采用python的zerorpc进行解耦。

   使用时，先运行一个win32版本的rpc服务器端，然后在任意环境下运行cn_zipline，通过rpc请求进行交易。

3.总结：

三种方案：

1. windows python2.7 32 bit 环境，直接使用cn_zipline
2. windows pyhton2.7 32 bit 的trade.dll rpc服务器端， cn_zipline可以使用任意python环境。
3. python rpc的替代解决方案：TdxTradeServer项目提供了一个交易服务器，使用的也是trade.dll，是直接用c++开发的，相对于python混编之后的rpc来讲，性能应该会好一点，因为配置起来有点麻烦，所以我并没有采用这个。

1. bug修复推相应分支，哪个分支的bug推哪个（包含master）
2. 大功能的增加，新建相应的独立分支，例如asyncio版本接口，就新建asyncio分支。
3. 小功能的增加或者修改， 推dev分支。
4. dev分支累计较多功能且稳定后，合并到master分支。大功能独立分支稳定后，合并到master分支。

使用tdx数据bundle
回测示例的buy hold策略
买入 000001
起始日为2012-1-1

报如下错误

AttributeError: 'NoneType' object has no attribute 'index'

Debug跟了下，好像是1-16日的交易日历时间为 1-16 0.0 Nat NaT
这是数据导入出了问题，还是这个版本的交易日历有问题，还是那天停牌而系统处理失当，还是其他问题？
buy1.zip
附件为完整代码

跟踪时 到 tradesimulation.py 这个文件时
查看 [dt for dt,action in self.clock] 的结果为
[Timestamp('2012-01-04 00:00:00+0000', tz='UTC'), Timestamp('2012-01-04 13:45:00+0000', tz='UTC'), Timestamp('2012-01-04 21:00:00+0000', tz='UTC'), Timestamp('2012-01-04 21:00:00+0000', tz='UTC'), Timestamp('2012-01-05 00:00:00+0000', tz='UTC')。。。Timestamp('2012-01-16 00:00:00+0000', tz='UTC'), NaT, Timestamp('2012-01-16 13:45:00+0000', tz='UTC'), NaT, Timestamp('2012-01-17 00:00:00+0000', tz='UTC'), Timestamp('2012-01-17 13:45:00+0000', tz='UTC'), Timestamp('2012-01-17 21:00:00+0000', tz='UTC'), Timestamp('2012-01-17 21:00:00+0000', tz='UTC'),

又跟踪了下
在 zipline/algorithm.py(531)_create_clock()
529 """
530 trading_o_and_c = self.trading_calendar.schedule.ix[
531 self.sim_params.sessions]
--> 532 market_closes = trading_o_and_c['market_close']

这传出的close时间就已经有NAT值了，这个值是何用意？
2012-01-13 00:00:00+00:00 2012-01-13 21:00:00
2012-01-16 00:00:00+00:00 NaT
2012-01-17 00:00:00+00:00 2012-01-17 21:00:00

2012-02-17 00:00:00+00:00 2012-02-17 21:00:00
2012-02-20 00:00:00+00:00 NaT
2012-02-21 00:00:00+00:00 2012-02-21 21:00:00

问题1：
config.json中dll路径配置正确，却报错无法找到指定模块

解决方案：
可能因缺少trade.dll的依赖dll未安装，使用depends程序查看trade.dll的依赖，并安装缺失dll。例如，题主win7运行正常，win10因缺少mfc120.dll而报错。安装ms visual c++ 2013 redistributable之后解决。

version: 1.0.0.16, python3.5
----command:
zipline ingest -b tdx --start 20090101 --overwrite False --minute False

----error message:
File "d:\toolset\anaconda3\envs\py35\lib\site-packages\zipline\data\bundles\tdx_bundle.py", line 52, in fetch_single_equity
df = engine.get_security_bars(symbol, freq, start, end)
File "d:\toolset\anaconda3\envs\py35\lib\site-packages\tdx\engine.py", line 228, in get_security_bars
['year', 'month', 'day', 'hour', 'minute'], axis=1)
File "d:\toolset\anaconda3\envs\py35\lib\site-packages\pandas\core\generic.py", line 1877, in drop
new_axis = axis.drop(labels, errors=errors)
File "d:\toolset\anaconda3\envs\py35\lib\site-packages\pandas\indexes\base.py", line 3051, in drop
labels[mask])
ValueError: labels ['year' 'month' 'day' 'hour' 'minute'] not contained in axis

当ingest 数据时提示sid错误，具体描述如下：

• Operating System: mac osx
• Python Version: python 3.5

(zipline_chn_py3.5) bogon:zipline administrator$ zipline ingest -b tdx --minute False --start 20170601
Merging daily equity files: [####################################] 100% 600680 [2017-07-10 05:31:36.981322] WARNING: engine: no k line data for 601869
[2017-07-10 05:31:36.995383] WARNING: engine: no k line data for 603713
Merging daily equity files: [####################################] 100%
/Users/administrator/Application/anaconda3/envs/zipline_chn_py3.5/lib/python3.5/site-packages/zipline/data/us_equity_pricing.py:958: FutureWarning: Conversion of the second argument of issubdtype from float to np.floating is deprecated. In future, it will be treated as np.float64 == np.dtype(float).type.
if not issubdtype(actual, expected):
[2017-07-10 05:33:29.916303] WARNING: UsEquityPricing: Couldn't compute ratio for dividend {'ex_date': numpy.datetime64('2017-07-21T00:00:00.000000000'), 'sid': 1, 'amount': 0.15800000429153443}
[2017-07-10 05:33:29.916535] WARNING: UsEquityPricing: Couldn't compute ratio for dividend {'ex_date': numpy.datetime64('2017-08-29T00:00:00.000000000'), 'sid': 2, 'amount': 0.7900000095367432}
[2017-07-10 05:33:29.916940] WARNING: UsEquityPricing: Couldn't compute ratio for dividend {'ex_date': numpy.datetime64('2018-06-22T00:00:00.000000000'), 'sid': 6, 'amount': 0.1809999942779541}
[2017-07-10 05:33:29.917154] WARNING: UsEquityPricing: Couldn't compute ratio for dividend {'ex_date': numpy.datetime64('2018-06-07T00:00:00.000000000'), 'sid': 8, 'amount': 0.05}
[2017-07-10 05:33:29.917301] WARNING: UsEquityPricing: Couldn't compute ratio for dividend {'ex_date': numpy.datetime64('2017-08-16T00:00:00.000000000'), 'sid': 9, 'amount': 0.020000000298023225}
[2017-07-10 05:33:29.917461] WARNING: UsEquityPricing: Couldn't compute ratio for dividend {'ex_date': numpy.datetime64('2018-06-29T00:00:00.000000000'), 'sid': 11, 'amount': 0.3}
[2017-07-10 05:33:29.917619] WARNING: UsEquityPricing: Couldn't compute ratio for dividend {'ex_date': numpy.datetime64('2017-07-19T00:00:00.000000000'), 'sid': 12, 'amount': 0.1}
[2017-07-10 05:33:29.917783] WARNING: UsEquityPricing: Couldn't compute ratio for dividend {'ex_date': numpy.datetime64('2018-06-27T00:00:00.000000000'), 'sid': 12, 'amount': 0.05}
[2017-07-10 05:33:29.917957] WARNING: UsEquityPricing: Couldn't compute ratio for dividend {'ex_date': numpy.datetime64('2018-06-22T00:00:00.000000000'), 'sid': 14, 'amount': 0.010000000149011612}
[2017-07-10 05:33:29.918090] WARNING: UsEquityPricing: Couldn't compute ratio for dividend {'ex_date': numpy.datetime64('2018-06-06T00:00:00.000000000'), 'sid': 16, 'amount': 0.16200000047683716}
[2017-07-10 05:33:29.918221] WARNING: UsEquityPricing: Couldn't compute ratio for dividend {'ex_date': numpy.datetime64('2017-07-14T00:00:00.000000000'), 'sid': 18, 'amount': 0.06000000238418579}
[2017-07-10 05:33:29.918403] WARNING: UsEquityPricing: Couldn't compute ratio for dividend {'ex_date': numpy.datetime64('2018-06-28T00:00:00.000000000'), 'sid': 21, 'amount': 0.05}
[2017-07-10 05:33:29.918534] WARNING: UsEquityPricing: Couldn't compute ratio for dividend {'ex_date': numpy.datetime64('2017-08-01T00:00:00.000000000'), 'sid': 22, 'amount': 0.4960000038146973}
[2017-07-10 05:33:29.918662] WARNING: UsEquityPricing: Couldn't compute ratio for dividend {'ex_date': numpy.datetime64('2018-05-23T00:00:00.000000000'), 'sid': 22, 'amount': 1.31899995803833}
[2017-07-10 05:33:29.918795] WARNING: UsEquityPricing: Couldn't compute ratio for dividend {'ex_date': numpy.datetime64('2017-08-15T00:00:00.000000000'), 'sid': 23, 'amount': 0.04000000059604645}
[2017-07-10 05:33:29.918976] WARNING: UsEquityPricing: Couldn't compute ratio for dividend {'ex_date': numpy.datetime64('2018-06-22T00:00:00.000000000'), 'sid': 27, 'amount': 0.0800000011920929}
[2017-07-10 05:33:29.919131] WARNING: UsEquityPricing: Couldn't compute ratio for dividend {'ex_date': numpy.datetime64('2018-06-13T00:00:00.000000000'), 'sid': 28, 'amount': 0.3}
[2017-07-10 05:33:29.919308] WARNING: UsEquityPricing: Couldn't compute ratio for dividend {'ex_date': numpy.datetime64('2018-07-10T00:00:00.000000000'), 'sid': 31, 'amount': 0.0550000011920929}
Traceback (most recent call last):
File "/Users/administrator/Application/anaconda3/envs/zipline_chn_py3.5/bin/zipline", line 11, in
sys.exit(main())
File "/Users/administrator/Application/anaconda3/envs/zipline_chn_py3.5/lib/python3.5/site-packages/click/core.py", line 722, in call
return self.main(*args, **kwargs)
File "/Users/administrator/Application/anaconda3/envs/zipline_chn_py3.5/lib/python3.5/site-packages/click/core.py", line 697, in main
rv = self.invoke(ctx)
File "/Users/administrator/Application/anaconda3/envs/zipline_chn_py3.5/lib/python3.5/site-packages/click/core.py", line 1066, in invoke
return _process_result(sub_ctx.command.invoke(sub_ctx))
File "/Users/administrator/Application/anaconda3/envs/zipline_chn_py3.5/lib/python3.5/site-packages/click/core.py", line 895, in invoke
return ctx.invoke(self.callback, **ctx.params)
File "/Users/administrator/Application/anaconda3/envs/zipline_chn_py3.5/lib/python3.5/site-packages/click/core.py", line 535, in invoke
return callback(*args, **kwargs)
File "/Users/administrator/Application/anaconda3/envs/zipline_chn_py3.5/lib/python3.5/site-packages/zipline/main.py", line 440, in ingest
writer=writer
File "/Users/administrator/Application/anaconda3/envs/zipline_chn_py3.5/lib/python3.5/site-packages/zipline/data/bundles/core.py", line 503, in ingest
pth.data_path([name, timestr], environ=environ),
File "/Users/administrator/Application/anaconda3/envs/zipline_chn_py3.5/lib/python3.5/site-packages/zipline/data/bundles/tdx_bundle.py", line 326, in tdx_bundle
shares=shares
File "/Users/administrator/Application/anaconda3/envs/zipline_chn_py3.5/lib/python3.5/site-packages/zipline/data/us_equity_pricing.py", line 1239, in write
self.write_dividend_data(dividends, stock_dividends)
File "/Users/administrator/Application/anaconda3/envs/zipline_chn_py3.5/lib/python3.5/site-packages/zipline/data/us_equity_pricing.py", line 1149, in write_dividend_data
dividend_ratios = self.calc_dividend_ratios(dividends)
File "/Users/administrator/Application/anaconda3/envs/zipline_chn_py3.5/lib/python3.5/site-packages/zipline/data/us_equity_pricing.py", line 1072, in calc_dividend_ratios
sid, prev_close_date, 'close')
File "/Users/administrator/Application/anaconda3/envs/zipline_chn_py3.5/lib/python3.5/site-packages/zipline/data/us_equity_pricing.py", line 766, in get_value
ix = self.sid_day_index(sid, dt)
File "/Users/administrator/Application/anaconda3/envs/zipline_chn_py3.5/lib/python3.5/site-packages/zipline/data/us_equity_pricing.py", line 734, in sid_day_index
offset = day_loc - self._calendar_offsets[sid]
KeyError: 34

首先数据有几部分：

1. 1分钟k线
2. 日K线
3. 6大表，报表基本面数据
4. 除息除权，股票基本信息， 股本数据

基本情况如下：

1. 1分钟k线，每只股票一个bcolz存放，可以增量
2. 日线数据是一个bcolz表放所有的股票日线数据，一个个股票按照顺序存放，没办法增量
3. 6大表基本面数据，sqlite可以增量
   4 没必要增量，很快，

我想在zipline中加入基本面选股的初衷是，ricequant本身提供了基本面数据，国外的量化发展比较成熟，Quantopian的zipline也可以通过一些付费的接口直接获取基本面数据。

目前的话，打算用tushare的基本面数据。但是有人说tushare的基本面数据不准确不太敢用。

但是如果通过wind等付费渠道获取数据的话，一方面wind的接口不是很好用，我目前还没有找到季报发布日期这个数据的获取方式。其它渠道的话，我也不是很清楚。

并且，付费渠道就需要维护一份数据供大家下载，而且要时常更新，十分的不方便。最好的方式还是通过接口集成到平台上。

不知道你们有什么好主意。