我正在尝试使用线性回归添加平滑趋势线，请帮助我获得时间序列数据

Question

ggplot()+
  geom_line(data=combined,aes(x=reorder(dates,value),y=value,color=variable,group=variable))+
  scale_y_continuous(labels=function(x) format(x,scientific=FALSE))+
  theme_gray()+
  theme(axis.text.x = element_text(angle=90),
        plot.title=element_text(hjust=0.5),plot.subtitle = 
          element_text(hjust = 0.5))+
  annotate(geom = 'text',x='03.5.20',y=150000,label=
             'WHO declares Covid-19 a Pandemic')+
  annotate(geom = 'point',x='03.11.20',y=125865,size=6,shape=21,fill='blue')+
  labs(title='Cases in China vs World ',x='Daily trend from January to March',y='Case Numbers ',
       subtitle = 'Data From Jan 22,2020 - Mar 23,2020')

这是我的常规 table，现在我正在尝试使用线性回归添加平滑趋势线。我尝试使用 stat_smooth(method='lm',formula=?) 但我正在使用的示例使用 y-x.

我的问题是在我的 x 轴上，我有日期，我不确定从这里去哪里。

这是我正在使用的数据

dates variable value
1  01.22.20    World   555
2  01.23.20    World   653
3  01.24.20    World   941
4  01.25.20    World  1434
5  01.26.20    World  2118
6  01.27.20    World  2927
7  01.28.20    World  5578
8  01.29.20    World  6166
9  01.30.20    World  8234
10 01.31.20    World  9927
63  01.22.20    China    548
64  01.23.20    China    643
65  01.24.20    China    920
66  01.25.20    China   1406
67  01.26.20    China   2075
68  01.27.20    China   2877
69  01.28.20    China   5509
70  01.29.20    China   6087
71  01.30.20    China   8141
72  01.31.20    China   9802

任何有关如何处理此问题的提示都将不胜感激。

Answer 1

你可以使用geom_smooth，如果你不想在平滑中使用颜色你可以将美学设置为geom_line

数据

df <- structure(list(dates = structure(c(18283, 18284, 18285, 18286, 
18287, 18288, 18289, 18290, 18291, 18292, 18283, 18284, 18285, 
18286, 18287, 18288, 18289, 18290, 18291, 18292), class = "Date"), 
    variable = c("World", "World", "World", "World", "World", 
    "World", "World", "World", "World", "World", "China", "China", 
    "China", "China", "China", "China", "China", "China", "China", 
    "China"), value = c(555L, 653L, 941L, 1434L, 2118L, 2927L, 
    5578L, 6166L, 8234L, 9927L, 548L, 643L, 920L, 1406L, 2075L, 
    2877L, 5509L, 6087L, 8141L, 9802L)), class = "data.frame", row.names = c(NA, 
-20L))

代码

df %>% 
  ggplot(aes(x=reorder(dates,value),y=value,color=variable,group=variable))+
  geom_line()+
  geom_smooth(method = "lm", se = FALSE)

一般顺利

df %>% 
  ggplot(aes(x=reorder(dates,value),y=value))+
  geom_line(aes(color=variable,group=variable))+
  geom_smooth(aes(group = 1),method = "lm", se = FALSE)

Answer 2

对于不重叠的日期，

df <- df %>%
  mutate(dates = as.Date(dates, format = "%m.%d.%y")) 
df %>%
  ggplot(aes(x = dates, y = value, group = variable, color = variable)) +
  geom_line() +
  geom_smooth(data = subset(df, variable == "China"), aes( dates, value), method = "lm", se = FALSE) +
  scale_y_continuous(labels=function(x) format(x,scientific=FALSE))+
  theme_gray()+
  theme(plot.title=element_text(hjust=0.5),plot.subtitle = 
        element_text(hjust = 0.5)) + 
  annotate(geom = 'text',x= as.Date('03.5.20', "%m.%d.%y"),y=150000,label=
             'WHO declares Covid-19 a Pandemic')+
  annotate(geom = 'point',x=as.Date('03.11.20', "%m.%d.%y"),y=125865,size=6,shape=21,fill='blue')+
  labs(title='Cases in China vs World ',x='Daily trend from January to March',y='Case Numbers ',
       subtitle = 'Data From Jan 22,2020 - Mar 23,2020')

我正在尝试使用线性回归添加平滑趋势线，请帮助我获得时间序列数据

I am trying to add a smooth trend line using linear regression, Help me i have time series data

r

data-visualization

time-series

linear-regression

trendline

数据

代码

一般顺利