如何获取多列时间序列数据框中的每小时平均值

Question

I have a time series data with 3 columns with Dates,energy values and Station names. I want to obtain the hourly average of the energy values separately for each station.

My data looks like this

df

     Datetime          Energy  Station
1 2016-01-01 07:19:00 743.0253   Ajmer
2 2016-01-01 07:20:00 765.7225   Ajmer
3 2016-01-01 07:21:00 788.1493   Ajmer
4 2016-01-01 08:20:00 834.7815   Ajmer
5 2016-01-01 08:21:00 857.3012   Ajmer
6 2016-01-31 16:58:00 3427.098  Kotada
7 2016-01-31 16:59:00 3397.591  Kotada
8 2016-01-31 17:00:00 3344.149  Kotada
9 2016-01-31 17:01:00 3270.803  Kotada

Expected Output:

     Datetime          Energy    Station
1. 2016-01-01 07:00:00 765.6324   Ajmer
2. 2016-01-01 08:00:00 846.0413   Ajmer
3. 2016-01-01 16:00:00 3412.345   Kotada
4. 2016-01-01 17:00:00 3307.476   Kotada

I tried group_by function to form a grouped data frame by Station names and then use the aggregate function to obtain the hourly average. But its not working.

> byStn=df %>% group_by(Station)
> hour_byStn=byStn %>% 
+            aggregate(energy,                                      
+                       list(hourtime = cut(Datetime, breaks="hour")),  
+                       mean, na.rm = TRUE)

I obtained the following error : Error in cut(Datetime, breaks = "hour") : object 'Datetime' not found.

Can you please tell me how to do this. This is the first time I am working with time series data and dpylr package as well.

Answer 1

We can use floor_date from lubridate to floor the 'DateTime' by hour ly interval, use that in group_by along with 'Station' and get the mean of 'Energy'

library(lubridate)
library(tidyverse)
df %>%
    group_by(Datetime = floor_date(Datetime, "hour"), Station) %>%
    summarise(Energy = mean(Energy, na.rm = TRUE))
# A tibble: 4 x 3
# Groups:   Datetime [4]
#  Datetime            Station Energy
#  <dttm>              <chr>    <dbl>
#1 2016-01-01 07:00:00 Ajmer     766.
#2 2016-01-01 08:00:00 Ajmer     846.
#3 2016-01-31 16:00:00 Kotada   3412.
#4 2016-01-31 17:00:00 Kotada   3307.

data

df <- structure(list(Datetime = structure(c(1451650740, 1451650800, 
1451650860, 1451654400, 1451654460, 1454277480, 1454277540, 1454277600, 
1454277660), class = c("POSIXct", "POSIXt"), tzone = ""), Energy = c(743.0253, 
765.7225, 788.1493, 834.7815, 857.3012, 3427.098, 3397.591, 3344.149, 
3270.803), Station = c("Ajmer", "Ajmer", "Ajmer", "Ajmer", "Ajmer", 
"Kotada", "Kotada", "Kotada", "Kotada")), row.names = c("1", 
"2", "3", "4", "5", "6", "7", "8", "9"), class = "data.frame")

Answer 2

I haven't tested it but you want something along the lines of this...

df %>%
    mutate(hourtime = cut(Datetime, breaks='hour')) %>%
    group_by(Station, hourtime) %>%
    summarise(avg_energy = mean(Energy, na.rm = T))

I would suggest maybe reading up on some basic dplyr syntax. I referenced this religiously when I first started using it: https://cran.r-project.org/web/packages/dplyr/vignettes/dplyr.html