数据分析之《菜鸟侦探挑战数据分析》-6-r语言-时间序列-逻辑回归分析-自相关-日历图

1、绘制时间序列图

# 加载小太郎家的失窃数额数据
xiaotailang <- read.csv("xiaotailang.csv", colClasses = c("numeric", "factor", "Date", "factor", "factor"))

# 安装便于处理时间序列数据的程序包
install.packages("xts")
library(xts)

# 给数据设置日期
library(dplyr)
lostD <- seq(as.Date("2014-10-1"), as.Date("2015-5-30"), by = "days")
lost <- xts(xiaotailang$数额, order.by = lostD, dateFormat = "Date")

lost %>% plot(type = "l", main = "失窃数额")

plot(lost[1:31], type = "l",main = "失窃数额（10月）", family = "sans")

# 绘制自相关图分析周期性
# 判断数据是否以2周为1个周期
lost %>% acf
从图中可以看出自相关第14的系数比较大