CODEWITHSUNDEEP
rtimemeanstandard-error
Greeny
Greeny

Reputation: 7

Calculating means and standard errors for time data

Is there a way to calculate means and standard error for time duration data? I have tried to work through similar fixes on this site but have not been successful.

The data needs to be grouped by "Site" and "Treatment".

I'd like to be able to calculate the mean and standard error of "HH.MM.SS' for each "Treatment" according to each "Site".

    names(Socials)
    [1] "Site"      "Date"      "Treatment" "Block"     "HH.MM.SS"  "Max.N"

A sample of the data is below (one of two sites):

    Site       Date Treatment Block             HH.MM.SS Max.N
    MH 2018-10-06         C     1     2020-07-05 00:08:05           9
    MH 2018-10-08         D     1     2020-07-05 00:12:56           6
    MH 2018-10-07         V     1     2020-07-05 00:57:52           4
    MH 2018-10-05     V + D     1     2020-07-05 00:53:14           4
    MH 2018-10-12         C     2     2020-07-05 00:03:57           7
    MH 2018-10-10         D     2     2020-07-05 00:21:34           2
    MH 2018-10-11         V     2     2020-07-05 01:10:24           4
    MH 2018-10-09     V + D     2     2020-07-05 01:15:17           6
    MH 2018-10-15         C     3     2020-07-05 00:08:44           1
    MH 2018-10-13         D     3     2020-07-05 00:58:22          23
    MH 2018-10-14         V     3     2020-07-05 00:00:07           3
    MH 2018-10-16     V + D     3     2020-07-05 01:09:12           4


    # CreatePOSIXct time objects for date.
    Socials$Date <- as.POSIXct(Socials$Date, format = "%d/%m/%y")

    # CreatePOSIXct time objects for the total time duration for each treatment
    Socials$HH.MM.SS <- as.POSIXct(Socials$HH.MM.SS, format = "%H:%M:%OS")

I have not provided all the data here, so the data could be subset, as per the below, to produce a working example

   #Subset Data - split sites MH and GI 
    MH.data <- subset(Socials, Site == 'MH')
    GI.data <- subset(Socials, Site == 'GI')

    ##### Calculate summary statistics for Observations ######
    library(data.table) 
    library(plotrix)

    df1 = data.table(Socials)

    df1_output = df1[, .("HH.MM.SS" = mean(HH.MM.SS), 
                   "std" = std.error(Treatment)), 
               by = c("Site", "Treatment")]

Any assistance would be greatly appreciated.

Thanks in advance. I hope the information provided here is okay.

Upvotes: 0

Views: 120

Answers (1)

Derek Fulton
Derek Fulton

Reputation: 328

I think you can just do it normally. This just worked for me:

> x=c(as.Date("2018-10-06"), as.Date("2018-11-06"), as.Date("2018-12-06"))
> mean(x)
[1] "2018-11-05"
> sd(x)
[1] 30.50137
> std.error(x)
[1] 17.60997

Upvotes: 0

Related Questions