利用JanusGraph图数据库统计股票涨停板后的走势
2024-06-09 05:04:51
环境信息：
1. Java 8
2. JanusGraph 0.54 （使用Oracle Berkeley DB Java Edition作为底层存储）
3. Windows 10
一、股票数据源
注册 Tushare https://tushare.pro/register?reg=332190 获取数据。
二、导入JanusGraph
1. 启动JanusGraph
打开cmd, 进入janusGraph根目录，输入
bin\gremlin-server.bat conf\gremlin-server\gremlin-server-berkeleyje.yaml2. 定义模型
打开另一个cmd, 进入janusGraph根目录，输入
bin\gremlin.bat
gremlin> :remote connect tinkerpop.server conf/remote.yaml
==>Configured localhost/127.0.0.1:8182
gremlin> :remote console
==>All scripts will now be sent to Gremlin Server - [localhost/127.0.0.1:8182] - type ':remote console' to return to local mode
复制黏贴以下代码建立模型。日K线即为图中的顶点，K线图是链表结构，组成有向无环图。
mgmt = graph.openManagement()
tsCode = mgmt.makePropertyKey('tsCode').dataType(String.class).cardinality(org.janusgraph.core.Cardinality.SINGLE).make()
tradeDate = mgmt.makePropertyKey('tradeDate').dataType(Date.class).cardinality(org.janusgraph.core.Cardinality.SINGLE).make()
open = mgmt.makePropertyKey('open').dataType(Double.class).cardinality(org.janusgraph.core.Cardinality.SINGLE).make()
high = mgmt.makePropertyKey('high').dataType(Double.class).cardinality(org.janusgraph.core.Cardinality.SINGLE).make()
low = mgmt.makePropertyKey('low').dataType(Double.class).cardinality(org.janusgraph.core.Cardinality.SINGLE).make()
close = mgmt.makePropertyKey('close').dataType(Double.class).cardinality(org.janusgraph.core.Cardinality.SINGLE).make()
preClose = mgmt.makePropertyKey('preClose').dataType(Double.class).cardinality(org.janusgraph.core.Cardinality.SINGLE).make()
ma5 = mgmt.makePropertyKey('ma5').dataType(Double.class).cardinality(org.janusgraph.core.Cardinality.SINGLE).make()
ma10 = mgmt.makePropertyKey('ma10').dataType(Double.class).cardinality(org.janusgraph.core.Cardinality.SINGLE).make()
ma20 = mgmt.makePropertyKey('ma20').dataType(Double.class).cardinality(org.janusgraph.core.Cardinality.SINGLE).make()
ma30 = mgmt.makePropertyKey('ma30').dataType(Double.class).cardinality(org.janusgraph.core.Cardinality.SINGLE).make()
ma90 = mgmt.makePropertyKey('ma90').dataType(Double.class).cardinality(org.janusgraph.core.Cardinality.SINGLE).make()
ma120 = mgmt.makePropertyKey('ma120').dataType(Double.class).cardinality(org.janusgraph.core.Cardinality.SINGLE).make()
ma250 = mgmt.makePropertyKey('ma250').dataType(Double.class).cardinality(org.janusgraph.core.Cardinality.SINGLE).make()
change = mgmt.makePropertyKey('change').dataType(Double.class).cardinality(org.janusgraph.core.Cardinality.SINGLE).make()
pctChange = mgmt.makePropertyKey('pctChange').dataType(Double.class).cardinality(org.janusgraph.core.Cardinality.SINGLE).make()
volume = mgmt.makePropertyKey('volume').dataType(Double.class).cardinality(org.janusgraph.core.Cardinality.SINGLE).make()
amount = mgmt.makePropertyKey('amount').dataType(Double.class).cardinality(org.janusgraph.core.Cardinality.SINGLE).make()
tradeRecord = mgmt.makeVertexLabel('tradeRecord').make()
mgmt.addProperties(tradeRecord, tsCode, tradeDate, open, high, low, close, preClose, ma5, ma10, ma20, ma30, ma90, ma120, ma250, change, pctChange, volume, amount)
mgmt.commit()3. 导入数据
导入数据各有各法，实现每支股票的K线 k0 -> k1 -> k2 -> ... -> kn
GraphTraversalSource g = traversal().withRemote("conf/remote-graph.properties");// 前一个日k线
Vertex kp = null;
// 当前日k线
Vertex kc;for (...) {String tsCode = ...;Date tradeDate = ...;Double open = ...;Double high = ...;Double low = ...;Double close = ...;Double ma5 = ...;Double ma10 = ...;Double ma20 = ...;Double ma30 = ...;Double ma90 = ...;Double ma120 = ...;Double ma250 = ...;Double change = ...;Double pctChange = ...;Double volume = ...;Double amount = ...;kc = g.addV("tradeRecord").property("tsCode", tsCode).property("tradeDate", tradeDate).property("open", open).property("high", high).property("low", low).property("close", close).property("ma5", ma5).property("ma10", ma10).property("ma20", ma20).property("ma30", ma30).property("ma90", ma90).property("ma120", ma120).property("ma250", ma250).property("change", change).property("pctChange", pctChange).property("volume", volume).property("amount", amount).next();if (null != kp) {g.addE("links").from(kp).to(kc).next();}kp = kc;
}
三、找出涨停板后未来两个交易日走势
try (GraphTraversalSource g = traversal().withRemote("conf/remote-graph.properties")) {GraphTraversal<Vertex, Map<String, Object>> t = g.with(Tokens.ARGS_EVAL_TIMEOUT, 3000000L).V().and(has("tsCode", "000001.SZ"), has("pctChange", P.gte(9.99))).as("x") // 涨停板.out().as("y") // 涨停板后第一个K线.out().as("z") // 涨停板后第二个K线.project("d1", "d2", "d2pctChange", "d3", "d3pctChange").by(select("x").by("tradeDate")) // 涨停板日期.by(select("y").by("tradeDate")) // 涨停板后第一个K线日期.by(select("y").by("pctChange")) // 涨停板后第一个K线涨跌幅.by(select("z").by("tradeDate")) // 涨停板后第二个K线日期.by(select("z").by("pctChange")) // 涨停板后第二个K线涨跌幅.select("d1", "d2", "d2pctChange", "d3", "d3pctChange").order().by("d1");while (t.hasNext()) {Map<String, Object> e = t.next();String r = String.format("%s,%s,%s,%s,%s \n",sdf.format(e.get("d1")), sdf.format(e.get("d2")), e.get("d2pctChange"), sdf.format(e.get("d3")), e.get("d3pctChange"));System.out.print(r);}} catch (Exception e) {e.printStackTrace();}输出结果：
000001.SZ,19921001,19921130,19921201,2.35,19921202,-1.53
000001.SZ,19940228,19940429,19940503,12.57,19940504,-8.06
000001.SZ,19940304,19940503,19940504,-8.06,19940505,3.63
000001.SZ,19940602,19940801,19940802,2.08,19940803,10.2
000001.SZ,19940604,19940803,19940804,1.76,19940805,4.19
000001.SZ,19940707,19940905,19940906,5.16,19940907,-8.98
000001.SZ,19941016,19941215,19941216,-1.59,19941219,-4.41
000001.SZ,19950319,19950518,19950519,2.99,19950522,1.58
000001.SZ,19960226,19960426,19960429,13.73,19960430,-2.52
000001.SZ,19960229,19960429,19960430,-2.52,19960502,-6.45
000001.SZ,19960401,19960531,19960603,7.27,19960604,1.67
000001.SZ,19960412,19960611,19960612,0.53,19960613,0.93
000001.SZ,19961019,19961218,19961219,-1.74,19961220,-0.06
000001.SZ,19970123,19970324,19970325,3.39,19970326,1.51
000001.SZ,19970306,19970505,19970506,9.99,19970507,9.96
000001.SZ,19970307,19970506,19970507,9.96,19970508,-8.18
000001.SZ,19970308,19970507,19970508,-8.18,19970509,4.54
000001.SZ,19970321,19970520,19970521,-1.87,19970522,-8.19
000001.SZ,19970421,19970620,19970623,7.57,19970624,4.6
000001.SZ,19990327,19990526,19990527,9.99,19990531,10.03
000001.SZ,19990328,19990527,19990531,10.03,19990601,8.41
000001.SZ,19990401,19990531,19990601,8.41,19990602,0.75
000001.SZ,19990425,19990624,19990625,1.09,19990628,8.97
000001.SZ,19990521,19990720,19990721,8.75,19990722,-5.22
000001.SZ,19991216,20000214,20000215,-4.47,20000216,-3.24
000001.SZ,20010824,20011023,20011024,0.14,20011025,-2.34
000001.SZ,20020425,20020624,20020625,10.0,20020626,0.21
000001.SZ,20020426,20020625,20020626,0.21,20020627,2.29
000001.SZ,20021115,20030114,20030115,-2.1,20030116,1.52
000001.SZ,20030215,20030416,20030417,-2.25,20030418,-4.23
000001.SZ,20050206,20050407,20050408,10.05,20050411,2.21
000001.SZ,20050207,20050408,20050411,2.21,20050412,-4.32
000001.SZ,20050822,20051021,20051024,-1.42,20051025,-2.72
000001.SZ,20060225,20060426,20060427,4.55,20060428,7.21
000001.SZ,20080721,20080919,20080922,9.98,20080923,-6.22
000001.SZ,20080724,20080922,20080923,-6.22,20080924,-2.58
000001.SZ,20080814,20081013,20081014,-0.6,20081016,-5.99
000001.SZ,20081225,20090223,20090225,1.07,20090226,-8.91
000001.SZ,20090416,20090615,20090616,2.68,20090617,-1.73
000001.SZ,20090531,20090730,20090803,-1.26,20090804,-3.13
000001.SZ,20090903,20091102,20091103,0.85,20091104,1.84
000001.SZ,20130711,20130909,20130910,2.56,20130911,2.01
000001.SZ,20140929,20141128,20141201,-2.01,20141202,7.63
000001.SZ,20150209,20150410,20150413,1.1,20150414,-1.45
000001.SZ,20200507,20200706,20200707,-1.28,20200708,1.81
四、总结
纯粹个人娱乐，没有参考价值。
利用JanusGraph图数据库统计股票涨停板后的走势相关推荐
- 使用Python统计股票高开后的走势
股票高开是指当日股票的开盘价高于昨日的收盘价,为什么会这样?莫不是机构背着咱们搞交易? 股票开盘价由集合竞价产生,我们知道股市开盘时间为9:30-11:30~13:00-15:00,但集合竞价时间在9 ...
- 图数据库统计、趋势2023
近年来,随着企业和组织寻求更有效的方式来存储和处理数据,图形数据库变得越来越流行. 随着大数据的兴起,了解图形数据库的趋势和统计数据非常重要,尤其是涉及到图形数据库类别时. 在这篇博文中,我们将深入研 ...
- 为知识的海洋绘制航图 —— 利用CirroData-Graph图数据库构建知识图谱
人类对世界的认知日益广博.精深,如何将浩如烟海的知识有效组织和整理,方便高效存储和查询是一个重要问题.知识图谱(Knowledge Graph)的概念应运而生,简单的说知识图谱就是一个具有有向图结构的 ...
- 图解图库JanusGraph系列-解惑图数据库，你知道什么是图数据库吗？
解惑图数据库!你知道什么是图数据库吗? 大家好,我是洋仔,JanusGraph图解系列文章,实时更新~ 图数据库文章总目录: 整理所有图相关文章,请移步(超链):图数据库系列-文章总目录 地址:htt ...
- 解惑图数据库！你知道什么是图数据库吗？
简介 为什么需要图数据库? 设想一个场景:在金融的反欺诈场景下,当一个用户小李 请求订单,我们可以设定一个规则: 获取该用户的身份证号.注册手机号.银行预留手机号.银行卡号.紧急联系人等信息 通过这些 ...
- 如何在图数据库中训练图卷积网络模型
在图数据库中训练GCN模型,可以利用图数据库的分布式计算框架现实应用中大型图的可扩展解决方案 什么是图卷积网络? 典型的前馈神经网络将每个数据点的特征作为输入并输出预测.利用训练数据集中每个数据点的特 ...
- 图数据库应用系列（一）：金融智能风控
导语 图数据库是一种新型的NoSQL数据库管理系统,使用图论(GraphTheory)的"图"作为数据模型,将实体表示为图的节点,实体间的关系表示为图的边,能够直接揭示事物之间关联 ...
- 图数据库初探——1. 选择一个合适的图数据库
1. 图数据库概览 Tips: 不了解一个领域的时候,先去找找这个领域的白皮书. 想做一个知识图谱,需要调研一下图数据库,就直接去找这方面的白皮书.相对权威但是不是非常实时的内容 或者直接搜索 知识图 ...
- 图数据库JanusGraph实战[6]: JanusGraph+HBase+ElasticSearch的环境搭建
图数据库JanusGraph实战[6]: JanusGraph+HBase+ElasticSearch的环境搭建 作者:胡佳辉(家辉) 日期:2019年01月14日 CSDN博客:https://b ...
最新文章
- 纯CSS3制作的圆角效果按钮菜单
- 高效搭建Storm全然分布式集群
- Android内存管理机制详解,【安卓知识普及贴】之关于安卓内存管理机制的一些介绍和阐述。...
- 一个能够编写、运行SQL查询并可视化结果的Web应用：SqlPad
- 活动目录组策略统一管理桌面
- javascript总for of和for in的区别？
- Pyqt5：使用Qlabel标签进行视频播放
- Linux文件属性与管理
- html导航栏所有页面通用,在所有页面中包含HTML导航栏
- DM 关键字、屏蔽关键字
- spring源码深度解析系列——环境搭建丢失spring-cglib-repack-3.2.8.jar和spring-objenesis-repack-3.0.1.jar的解决办法
- 《大明王朝》掠之于官
- 3.抽象类（shape）
- 给AWStats关键词统计增加去百度，谷歌等搜索引擎的站内查询链接
- 微信小程序--wx-charts图表插件使用教程
- AU插件安装 - FabFilter、RX
- 计算机高级通信机制,深入电脑运行原理之进程通信（Operating System四级内容）...
- 学科实践活动感悟50字_包饺子实践活动感悟
- 一道和稀饭绝配的自制腌菜
- 【技术科普一】什么是石墨烯技术？