I have size data of different species. Each sample represents a plot of reef of 1 m^2 (a quadrat; "Unique"). There should be 5 quadrats at each site in any given year ("YrSi") and any number of species at a quadrat (some have more then others and they often differ). I need to calculate a mean of the "Size" based on the weighting of the "Count" (ie weighted mean) for each "YrSi"(year-site combo) and "Taxa" (ie species). Example:
head(df)
Unique Yr Si Qd YrSi Taxa Count Size SLength
6 2007-M1-1991 2007 M1 1991 2007 - M1 Carpophyllum flexuosum 7 10
7 2007-M1-1991 2007 M1 1991 2007 - M1 Carpophyllum flexuosum 1 15
8 2007-M1-1991 2007 M1 1991 2007 - M1 Carpophyllum flexuosum 5 20
9 2007-M1-1991 2007 M1 1991 2007 - M1 Carpophyllum flexuosum 4 25
10 2007-M1-1991 2007 M1 1991 2007 - M1 Carpophyllum flexuosum 4 30
11 2007-M1-1991 2007 M1 1991 2007 - M1 Carpophyllum flexuosum 1 35
I tried using weighted.mean embedded within ddply. but the calculation is wrong and I got the same value for all species in all YrSi. I suspect it applied the weighted.mean calculation across all species and samples.
wt_mean.df = ddply(df, c("YrSi","Taxa"),
function(x) weighted.mean(df$Size, df$Count))
head(wt_mean.df)
YrSi Taxa V1
1 2007 - C1 Buccinulum lineum 21.22346
2 2007 - C1 Cantharidus purpureus 21.22346
3 2007 - C1 Carpophyllum maschalocarpum 21.22346
4 2007 - C1 Cominella virgata 21.22346
5 2007 - C1 Cookia sulcata 21.22346
6 2007 - C1 Ecklonia radiata 21.22346
head(wt_mean.df)
YrSi Taxa V1
1 2007 - C1 Buccinulum lineum 21.22346
2 2007 - C1 Cantharidus purpureus 21.22346
3 2007 - C1 Carpophyllum maschalocarpum 21.22346
4 2007 - C1 Cominella virgata 21.22346
5 2007 - C1 Cookia sulcata 21.22346
6 2007 - C1 Ecklonia radiata 21.22346
The calculation is wrong and I got the same value for all species in all YrSi. I suspect it applied the weighted.mean calculation across all species and samples.
tail(wt_mean.df)
YrSi Taxa V1
1603 2019 - T5 Maoricolpus roseus 21.22346
1604 2019 - T5 Patiriella regularis 21.22346
1605 2019 - T5 Sargassum scabridum 21.22346
1606 2019 - T5 Sargassum sinclairii 21.22346
1607 2019 - T5 Trochus viridus 21.22346
1608 2019 - T5 Zonaria turneriana 21.22346
What am I doing wrong? why don't I get the correct weighted means in V1? Also, it would be good to also get a weighted sd, but I haven't looked into it yet. Please help.
Dplyr might be an easy solution for you.
library(dplyr)
output <- df%>%
group_by(Yr, Si, Taxa) %>%
summarise(wMean = weighted.mean(Size, Count))