{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#VN.DATAYES - Welcome!\n",
    "***\n",
    "##1. Preface\n",
    "\n",
    "###1.1\n",
    "vn.datayes是一个从属于vnpy的开源历史数据模块；使用通联数据API以及MongoDB进行数据的下载和存储管理。项目目前与将来主要解决\\准备解决以下问题：\n",
    "\n",
    "* 从通联数据等API高效地爬取、更新、清洗历史数据。\n",
    "* 基于MongoDB的数据库管理、快速查询、转换输出格式；支持自定义符合需求的行情历史数据库。\n",
    "* 基于Python.Matplotlib或R.ggplot2，快速绘制K线图等可视化对象。\n",
    "\n",
    "项目目前主要包括了通联API开发者试用方案中大部分的市场行情日线数据（股票、期货、期权、指数、基金等），以及部分基本面数据。数据下载与更新主要采用多线程设计，测试效率如下：\n",
    "\n",
    "| 数据集举例 | 数据集容量 | 下载时间估计  |\n",
    "| :-------------: | :-------------: | :-------------: |\n",
    "| 股票日线数据，2800个交易代码，2013年1月1日至2015年8月1日 | 2800个collection，约500条/each | 7-10分钟 |\n",
    "| 股票分钟线数据，2个交易代码，2013年1月1日至2015年8月1日 | 2个collection，约20万条/each      | 1-2分钟 |\n",
    "| 股票日线数据更新任务，2800个交易代码，2015年8月1日至2015年8月15日 | 2800个collection，约10条/each      | 1-2分钟 |\n",
    "\n",
    "vn.datayes基于MongoDB数据库，通过一个json配置文件简化数据库的初始化、设置、动态更新过程。较为精细的数据库操作仍需编写脚本进行。若对MongoDB与pymongo不熟悉，推荐使用Robomongo等窗口化查看工具作为辅助。\n",
    "\n",
    "###1.2 主要依赖：\n",
    "pymongo,  pandas,  requests,  json\n",
    "###1.3 开发测试环境：\n",
    "Mac OS X 10.10; Windows 7 || Anaconda.Python 2.7"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* * *\n",
    "##2. Get Started\n",
    "###2.1 准备\n",
    "\n",
    "* 下载并安装MongoDB: https://www.mongodb.org/downloads\n",
    "* 获取API token，以通联数据为例。\n",
    "\n",
    "![fig1](figs/fig1.png)\n",
    "\n",
    "* 更新pymongo至3.0以上版本; 更新requests等包。 \n",
    "```\n",
    "~$ pip install pymongo --upgrade\n",
    "~$ pip install requests --upgrade\n",
    "```\n",
    "\n",
    "* [ ! 注意，本模块需要pymongo3.0新加入的部分方法，使用vnpy本体所用的2.7版本对应方法将无法正常插入数据。依赖冲突的问题会尽快被解决，目前推荐制作一个virtual environment来单独运行这个模块；或者暴力切换pymongo的版本：]\n",
    "```\n",
    "~$ pip install pymongo==3.0.3  # this module.\n",
    "~$ pip install pymongo==2.7.2  # pymongo 2.7.\n",
    "```\n",
    "\n",
    "* 启动MongoDB\n",
    "```\n",
    "~$ mongod\n",
    "```\n",
    "\n",
    "\n",
    "###2.2 数据库初始化与下载\n",
    "* **api.Config** 对象包含了向API进行数据请求所需的信息，我们需要一个用户token来初始化这个对象。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'domain': 'api.wmcloud.com/data',\n",
       " 'header': {'Authorization': 'Bearer 7c2e59e212dbff90ffd6b382c7afb57bc987a99307d382b058af6748f591d723',\n",
       "  'Connection': 'keep-alive'},\n",
       " 'ssl': False,\n",
       " 'version': 'v1'}"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from storage import *\n",
    "\n",
    "myConfig = Config(head=\"Zed's Config\", \n",
    "                  token='7c2e59e212dbff90ffd6b382c7afb57bc987a99307d382b058af6748f591d723')\n",
    "myConfig.body"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* * *\n",
    "* **storage.DBConfig** 对象包含了数据库配置。我们需要自己编写一个json字典来填充这个对象。举例来说，我们希望下载股票日线数据和指数日线数据，数据库名称为DATAYES_EQUITY_D1和DATAYES_INDEX_D1，index为日期“date”。那么json字典是这样的："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'client': MongoClient('localhost', 27017),\n",
       " 'dbNames': ['EQU_M1', 'EQU_D1', 'FUT_D1', 'OPT_D1', 'FUD_D1', 'IDX_D1'],\n",
       " 'dbs': {'EQU_D1': {'collNames': 'equTicker',\n",
       "   'index': 'date',\n",
       "   'self': Database(MongoClient('localhost', 27017), u'DATAYES_EQUITY_D1')},\n",
       "  'EQU_M1': {'collNames': 'secID',\n",
       "   'index': 'dateTime',\n",
       "   'self': Database(MongoClient('localhost', 27017), u'DATAYES_EQUITY_M1')},\n",
       "  'FUD_D1': {'collNames': 'fudTicker',\n",
       "   'index': 'date',\n",
       "   'self': Database(MongoClient('localhost', 27017), u'DATAYES_FUND_D1')},\n",
       "  'FUT_D1': {'collNames': 'futTicker',\n",
       "   'index': 'date',\n",
       "   'self': Database(MongoClient('localhost', 27017), u'DATAYES_FUTURE_D1')},\n",
       "  'IDX_D1': {'collNames': 'idxTicker',\n",
       "   'index': 'date',\n",
       "   'self': Database(MongoClient('localhost', 27017), u'DATAYES_INDEX_D1')},\n",
       "  'OPT_D1': {'collNames': 'optTicker',\n",
       "   'index': 'date',\n",
       "   'self': Database(MongoClient('localhost', 27017), u'DATAYES_OPTION_D1')}}}"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "client = pymongo.MongoClient() # pymongo.connection object.\n",
    "\n",
    "body = {\n",
    "    'client': client, # connection object.\n",
    "    'dbs': {\n",
    "        'EQU_D1': {                              # in-python alias: 'EQU_D1'\n",
    "            'self': client['DATAYES_EQUITY_D1'], # pymongo.database[name] object.\n",
    "            'index': 'date',                     # index name.\n",
    "            'collNames': 'equTicker'             # what are collection names consist of.\n",
    "        },\n",
    "        'IDX_D1': {                              # Another database\n",
    "            'self': client['DATAYES_INDEX_D1'],\n",
    "            'index': 'date',\n",
    "            'collNames': 'idxTicker'\n",
    "        }\n",
    "    },\n",
    "    'dbNames': ['EQU_D1','IDX_D1']               # List of alias.\n",
    "}\n",
    "\n",
    "myDbConfig_ = DBConfig(body=body)\n",
    "\n",
    "# 这看上去有些麻烦；不想这么做的话可以直接使用DBConfig的默认构造函数。\n",
    "\n",
    "myDbConfig = DBConfig()\n",
    "\n",
    "myDbConfig.body"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* * *\n",
    "* **api.PyApi**是向网络数据源进行请求的主要对象。**storage.MongodController**是进行数据库管理的对象。当我们完成了配置对象的构造，即可初始化PyApi与MongodController。**MongodController._get_coll_names()** 和**MongodController._ensure_index()** 是数据库初始化所调用的方法，为了模块开发的方便，它们暂时没有被放进构造函数中自动执行。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[MONGOD]: Collection names gotten.\n",
      "[MONGOD]: MongoDB index set.\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "1"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "myApi = PyApi(myConfig) # construct PyApi object.\n",
    "mc = MongodController(api=myApi, config=myDbConfig) # construct MongodController object, \n",
    "                                                    # on the top of PyApi.\n",
    "mc._get_coll_names()    # get names of collections.\n",
    "mc._ensure_index()      # ensure collection indices."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![fig2](figs/fig2.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* 使用**MongodController.download#()**方法进行下载。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "mc.download_index_D1('20150101','20150801')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![fig3](figs/fig3.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###2.3 数据库更新\n",
    "* 使用**MongodController.update#()**方法进行更新。脚本会自动寻找数据库中的最后一日并更新至最新交易日。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "datetime.datetime(2015, 8, 17, 10, 49, 21, 37758)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from datetime import datetime\n",
    "datetime.now()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "mc.update_index_D1()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![fig4](figs/fig4.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "###2.4 Mac OS或Linux下的下载与更新\n",
    "模块中包含了一些shell脚本，方面在linux-like os下的数据下载、更新。\n",
    "```\n",
    "~$ cd path/of/vn/datayes\n",
    "~$ chmod +x prepare.sh\n",
    "~$ ./prepare.sh\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![fig5](figs/fig5.png)\n",
    "![fig6](figs/fig6.png)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}