Question

dput(df)
structure(list(Month = structure(c(15248, 15522), class = "Date"), 
    Value = c(1, 3)), .Names = c("Month", "Value"), row.names = 1:2, class = "data.frame")

ggplot(df, aes(Month, Value)) + 
  geom_bar(fill = "orange", size = .3, stat = "identity", position = "identity") +
  geom_smooth(data = df, aes(Month, Value, group = 1), method = "lm", 
              size = 2, color = "red") + 
  scale_x_date(breaks = "1 month", labels = date_format("%b-%Y"), 
               limits = as.Date(c('2011-01-01','2013-01-01')))

Bars in the ggplot goes over the other dates as well. I do get this warning message:

Warning message:
In qt((1 - level)/2, df) : NaNs produced

Is there a way to place the bin to the belonging date, rather than crossing over to other dates?

Answer 1

Simplified your code for plot. There is no need to write data frame name, x and y values in each geom.

To change bar width, argument width= in geom_bar() can be used.

ggplot(df, aes(Month, Value)) + 
  geom_bar(fill="orange",stat="identity",width=15)+
  geom_smooth( method="lm", size=2, color="red")+
  scale_x_date(breaks = "1 month", labels=date_format("%b-%Y"), limits = as.Date(c('2011-01-01','2013-01-01')))

You get error message because your data frame contains only two values (you can’t do regression with just two values). If there really are only two values, then replace

geom_smooth( method="lm", size=2, color="red")

with

geom_line(size=2, color="red")