Upload New File

策略5：融合技术指标的QP策略.ipynb

+{
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import numpy as np\n",
+    "import matplotlib.pyplot as plt\n",
+    "import cvxopt as opt\n",
+    "from cvxopt import blas, solvers\n",
+    "import pandas as pd\n",
+    "from cvxopt import solvers, matrix\n",
+    "\n",
+    "df=pd.read_csv('stocks.csv')\n",
+    "\n",
+    "stock_list=[row['stock'] for index,row in df.iterrows()]\n",
+    "\n",
+    "#选股池\n",
+    "instruments=stock_list\n",
+    "\n",
+    "#回测时间\n",
+    "start_date='2019-01-03'\n",
+    "end_date='2021-01-22'\n",
+    "\n",
+    "\n",
+    "def initialize(context):\n",
+    "    \n",
+    "    context.set_commission(PerOrder(buy_cost=0.0003,sell_cost=0.0003,min_cost=5))\n",
+    "\n",
+    "\n",
+    "    \n",
+    "\n",
+    "##每天执行一次\n",
+    "def handle_data(context,data):\n",
+    "    \n",
+    "    ##技术指标所需要的最大天数\n",
+    "    \n",
+    "    if context.trading_day_index<100:\n",
+    "        return\n",
+    "    \n",
+    "    if (context.trading_day_index)%20!=0:\n",
+    "        return \n",
+    "    \n",
+    "    print(context.trading_day_index)\n",
+    "    \n",
+    "    price_df=pd.DataFrame()\n",
+    "    \n",
+    "    stock_num=120\n",
+    "    count=0\n",
+    "    \n",
+    "    for stock in instruments:\n",
+    "        count+=1\n",
+    "        sid=context.symbol(stock)\n",
+    "        prices = data.history(sid,'price',5,'1d')\n",
+    "        price_df=pd.concat([price_df,prices],axis=1)\n",
+    "        price_df.fillna(0,inplace=True)\n",
+    "        if(count==stock_num):\n",
+    "            break\n",
+    "       \n",
+    "    ## 120,5\n",
+    "    price_np=price_df.to_numpy().T\n",
+    "   \n",
+    "    \n",
+    "    ##cov\n",
+    "    ##(120,120)\n",
+    "    cov_df= pd.DataFrame(np.cov(price_np))\n",
+    "    P=matrix(cov_df.to_numpy())\n",
+    "    \n",
+    "    q=-matrix(np.mean(price_np,axis=1),(stock_num,1))\n",
+    "    \n",
+    "    \n",
+    "    ##Ax=b\n",
+    "    A=matrix(1.,(1,stock_num))\n",
+    "    b=matrix(1.)\n",
+    "    \n",
+    "    \n",
+    "    ##GX<=h\n",
+    "    temp_df1= pd.DataFrame(-np.eye(stock_num))\n",
+    "    temp_df2= pd.DataFrame(np.eye(stock_num))\n",
+    "    G_df=pd.concat([temp_df1,temp_df2],axis=0)\n",
+    "    G = matrix(G_df.to_numpy()) \n",
+    "    \n",
+    "    ##xi<=0.2\n",
+    "    temp_list=[]\n",
+    "    for i in range(stock_num):\n",
+    "        temp_list.append([0.0])\n",
+    "        \n",
+    "    ##每只权重不超过0.2\n",
+    "    for i in range(stock_num):\n",
+    "        temp_list.append([0.2])\n",
+    "    h = opt.matrix(np.array(temp_list))\n",
+    "  \n",
+    "    \n",
+    "    ##solver:\n",
+    "    solvers.options['show_progress'] = False\n",
+    "    sol = solvers.qp(P,q,G,h,A,b)\n",
+    "    weight=list(sol['x'])\n",
+    "\n",
+    "    stock_selected_list=[]\n",
+    "    sum_weight=0.0\n",
+    "    cur=0\n",
+    "    for stock in stock_list:      \n",
+    "        sid=context.symbol(stock)\n",
+    "        if data.can_trade(sid):\n",
+    "            stock_selected_list.append(stock)\n",
+    "            sid=context.symbol(stock)\n",
+    "            \n",
+    "            ##融入技术指标，如果满足条件，，权重加倍\n",
+    "            ma_3=data.history(sid,'price',3,'1d').mean()\n",
+    "            ma_5=data.history(sid,'price',5,'1d').mean()\n",
+    "            ma_10=data.history(sid,'price',10,'1d').mean()\n",
+    "            ma_15=data.history(sid,'price',15,'1d').mean()\n",
+    "            sum_weight+=weight[cur]\n",
+    "            if ma_3>ma_5 and ma_5>ma_10 and ma_10>ma_15:\n",
+    "                    sum_weight+=weight[cur]*2\n",
+    "                    weight[cur]*=2\n",
+    "        else:\n",
+    "            weight[cur]=0\n",
+    "        cur+=1\n",
+    "    \n",
+    "    cur=0\n",
+    "    for stock in stock_list:\n",
+    "        if weight[cur]!=0:\n",
+    "            sid=context.symbol(stock)\n",
+    "            price=data.current(sid,'price')\n",
+    "            context.order_target_percent(sid,weight[cur]/sum_weight)\n",
+    "        cur+=1\n",
+    "\n",
+    "m=M.trade.v2(\n",
+    "    instruments=instruments,\n",
+    "    start_date=start_date,\n",
+    "    end_date=end_date,\n",
+    "    initialize=initialize,\n",
+    "    handle_data=handle_data,\n",
+    "    order_price_field_buy='open',\n",
+    "    order_price_field_sell='open',\n",
+    "    capital_base=1000000\n",
+    ")\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "is_mlstudio": "false",
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "name": ""
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 4
+}