在我的数据集中添加一个标准误差列，以便可以绘制误差线

Question

Data <- data.frame(id, consumption, Day, Hour)
#The data is a large time series data set with thousands of valued per household id.
#eg. 
consumption <- c(99, 119, 130, 110, 109, 118) etc.
#Hour and Day were calculated from the Date Time of the dataset.

I have created a two separate line graphs using ggplot2 for total mean energy consumption and mean energy consumption between 4 and 8pm for a range of households. I am looking to add value specific (not constant) error bars which correspond to the standard error for each value. I am unsure of how to add a standard error column to my data set corresponding to each individual value. If you could use pipes that would be great!

I have looked online for different methods to calculate individual standard errors and add a column with them however nothing has worked. It may be because I am not plotting the raw data, but instead plotting data that has been summarised (sum and mean). The two plots 1) and 2) will have different error bars for the same dates. I have included a picture to what the plot should look like at the end.

These are my plots: 1) Overall Daily Mean Consumption

Data %>%
 group_by(id, Day)%>% 
#id is household identification
  summarise(DailyCons = sum(consumption))%>% 
#Sum for total daily consumption per household
group_by(Day)%>%
 summarise(MeanDailyCons = mean(DailyCons))%>% 
#Find mean daily consumption for all households
  ggplot()+
  geom_line(aes(x= Day, y= MeanDailyCons))

2) Daily Mean between 16:00-20:00

Data %>%
  mutate(TimeInt = ifelse(Hour %in% c(16, 17, 18, 19, 20), Hour, NA))%>% 
#removing Hours outside of range 16-20
  group_by(id, TimeInt, Day) %>%
  na.omit(TimeInt)%>%
  summarise(sumPeakCons = sum(consumption)) %>% 
#sum for total consumption for each hour in interval for each house
  group_by(bmg_id, Day) %>%
  summarise(PeakCons = sum(sumPeakCons)) %>% 
#sum for total daily consumption in interval for each house
  group_by(Day) %>%
  summarise(DailyPeakCons = mean(PeakCons)) %>% 
# Daily mean consumption for all houses
  ggplot()+
  geom_line(aes(x= Day, y= DailyPeakCons))

An image is included to show the desired result.

https://i.stack.imgur.com/WDT8Z.png

Answer 1

You are correct that you cannot add the standard error after you summarize the data by day. Any function that would try would just receive a mean and a datetime, not enough to create an error. The standard error must be added when you summarize from the raw data.

Add another column to your summarise statement:

summarise(DailyPeakCons = mean(PeakCons),DailyPeakConsErr = sd(PeakCons)) %>%

This will give the standard deviation of each day's peak consumptions.