2025新策略宏观反欺诈算法.R

monthlyReturn()
quarterlyReturn()
weeklyReturn()
yearlyReturn()

start_date<-Sys.Date()-days(1460)



##获取股票的反弹数据
get_log_returns<-function(data){
  data%>%
    mutate(Log.Returns=dailyReturn(as.xts(as.numeric(Close),order.by=Date),subset=NULL,type='arithmetic',leading=TRUE),
           Month.Returns=monthlyReturn(as.xts(as.numeric(Close),order.by=Date),subset=NULL,type='arithmetic',leading=TRUE),
           Quarter.Returns=quarterlyReturn(as.xts(as.numeric(Close),order.by=Date),subset=NULL,type='arithmetic',leading=TRUE),
           Year.Returns=yearlyReturn(as.xts(as.numeric(Close),order.by=Date),subset=NULL,type='arithmetic',leading=TRUE))%>%
    select(Date,Log.Returns,Month.Returns,Quarter.Returns,Year.Returns)%>%
    as_tibble() }


##### 年线级别投资机会
get_log_returns<-function(data){
  data%>%
    mutate(Log.Returns=dailyReturn(as.xts(as.numeric(Close),order.by=Date),subset=NULL,type='arithmetic',leading=TRUE))%>%
           select(Date,Log.Returns)%>%
    as_tibble() }


stocks_name<-df_all%>%
  filter(industry_4 %in% c("港口","航运"))

## 2024年12月19号探索长城人寿的战略布局；
stocks_name<-df_all%>%
  filter(name %in% c("赣粤高速","五洲交通","上海能源","昊华能源","江南水务","城发环境","中原高速","楚天高速","淮河能源","创业环保","无锡银行","金融街",
"东莞控股","浙江交科","秦港股份","绿色动力"))

stocks_name
library(purrr)
china_stocks <- stocks_name%>%
  mutate(
    stock.prices = map(ts_code, 
                       function(.x) get_stock_prices(.x)
    ),
    log.returns  = map(stock.prices, 
                       function(.x) get_log_returns(.x)),
    mean.log.returns = map_dbl(log.returns, ~ mean(.$Log.Returns)),
    sd.log.returns   = map_dbl(log.returns, ~ sd(.$Log.Returns)),
    n.trade.days = map_dbl(stock.prices, nrow)
  )%>%as_tibble()

china_stocksdata<-china_stocks%>%
  unnest(stock.prices)%>%
  mutate(Time=as.numeric(today()-ymd(list_date)))

##################################   针对某个板块异常特定的走势图
china_stocks %>%
  unnest(stock.prices)%>%
  arrange(desc(Date))%>%
  ggplot(aes(x=desc(Date),y=as.numeric(Close)),color=name)+
  geom_line()+
  facet_wrap(~name,ncol=3,scales = "free_y")+
  labs(x="日期",y="价格",title="长城人寿中部崛起合纵与能源连横大战略")+
  theme_avatar(title.font = "Slayer",
               text.font = "Slayer",
               title.size = 14) 
  #theme_hildaDay(ticks = TRUE,
  #               legend.position = "none")

library(plotly)
#subplot出现了BUG
china_stocks %>%
  unnest(stock.prices)%>%
  select(Date,Close,name)%>%
  group_by(name) %>%
  do(p = plot_ly(., x = ~desc(ymd(Date)), y = ~Close,name=~name)%>%
       add_lines) %>%
  subplot(nrows =6, shareX = TRUE)%>%
  layout(title = "2019~2024年港口与航运集团股价走势")


china_stocks %>%
  unnest(stock.prices)%>%
  select(Date,Close,name)%>%
  group_by(name) %>%
  do(p = plot_ly(., x = ~desc(ymd(Date)), y = ~Close,name=~name)%>%
       add_lines) %>%
  subplot(nrows =6, shareX = TRUE)%>%
  layout(title = "2019~2024年港口与航运集团股价走势")

################################聚类算法
library(corrplot)
library(gplots)
library(RColorBrewer)
library(heatmaply)
library(d3heatmap)
library(purrr)
library(dplyr)
library(tidyverse)

####异常公司聚类
heatmapdata<-china_stocks%>%
  unnest(stock.prices)%>%
  select(Date,name,Close)%>%
  mutate(Close=as.numeric(Close))%>%
  spread(key=name, value=Close)%>%
  select(-Date)

d3heatmap(heatmapdata,scale = "column", 
          show_grid = TRUE, anim_duration = 500,
          distfun = dist, hclustfun = hclust,height = 750,width=1350,main="天网作战聚类算法可视化")%>% 
  hmLegend(show = TRUE, title = "物以类聚人以群分", location = "tl") 

#Heatmap3D(heatmapdata)


#############################################反欺诈算法过滤可疑对象
library(tidyr)
library(tidyverse)
library(timetk)
library(purrr)
library(dplyr)

anoma_data<-china_stocks %>%
  unnest(stock.prices)%>%
  select(Date,Volume,name)%>%
  group_by(name) %>%
  tk_anomaly_diagnostics(Date,Volume)%>%
  filter(Date>ymd(today()-10))

library(DT)
datatable(anoma_data)

head(anoma_data)
anoma_name<-anoma_data%>%filter(anomaly=="Yes")

anoma_data%>%filter(anomaly=="Yes")%>%select(name,anomaly,Date)%>%knitr::kable()

head(df_all)
###########异常交易数据
anoma_data%>%
  filter(anomaly=="Yes")%>%
  select(name,anomaly,Date)%>%
  left_join(df_all)%>%
  select(name,anomaly,Date,close,pct_chg)%>%
  arrange(desc(pct_chg))%>%
  knitr::kable()

anoma_namedata<-anoma_data%>%
  filter(anomaly=="Yes")%>%
  select(name,anomaly,Date)%>%
  left_join(df_all)%>%
  select(name,anomaly,Date,close,pct_chg)%>%
  arrange(desc(pct_chg))


#################################防欺诈算法异常数据表格

library(DT)
colnames(anoma_namedata)<-c('公司','是否异常','时间','收盘价','涨跌幅')
datatable(anoma_namedata,caption = htmltools::tags$caption(
  style = 'caption-side: top; text-align: center;',
  '反欺诈表: ', htmltools::em('主力异常行为数据')),extensions = 'Buttons', 
  options = list(dom = 'Bfrtip',pageLength = 20,
                 buttons = c('copy', 'csv', 'excel', 'pdf', 'print'),initComplete = JS(
                   "function(settings, json) {",
                   "$(this.api().table().header()).css({'background-color': 'lightpink', 'color': 'brown'});",
                   "$(this.api().table().body()).css({'background-color': 'gray', 'color': 'red'});",
                   "}"),searchHighlight = TRUE, filter = 'top'
  ))%>%
  formatStyle('涨跌幅',
              color = styleInterval(4, c('red', 'brown')),
              backgroundColor = styleInterval(2, c('yellow', 'lightblue')))%>%
  formatStyle(
    '是否异常',
    transform = 'rotateX(15deg) rotateY(1deg) rotateZ(1deg)',
    backgroundColor = styleEqual(
      unique(anoma_namedata$是否异常), c('lightgreen')
    ) )


#'background-color': '#000', 'color': '#fff'
plot_ly(anoma_namedata) %>% 
  add_table()%>%
  layout(title='反欺诈算法数据')



library(dplyr)
library(purrr)

library(timetk)

china_data<-china_stocks %>%
  unnest(stock.prices)
#####################################OK的
library(showtext)
showtext_auto(enable=TRUE)

a_plot1<-china_stocks %>%
  unnest(stock.prices)%>%
  select(Date,Volume,name)%>%
  filter(name %in% anoma_namedata$公司)%>%
  group_by(name) %>%
  plot_anomaly_diagnostics(desc(Date), Volume,
                           .message = FALSE,
                           .facet_ncol = 3,
                           .ribbon_alpha = 0.25,
                           .interactive = FALSE)
a_plot1

ggplotly(a_plot1)%>%layout(title = "天网作战地图之反欺诈算法")

colnames(anoma_data)<-c("名称","日期","观察次数","周期性","趋势性","提示信号","季节调整","一级水平","二级水平","异常","一级维度","二级维度")

datatable(anoma_data)