R help - May 2012 - weighted averages for two variables

Hi R users,

I have a dataset with multiple variables and i'm trying to weigh average
depths for fish species per year by their abundance (CPUE. I have tried the
weighted.mean function but as i have two columns for the x value the lenghts
differ with the w (CPUE). How do I solve this problem?

So far I have tried this:
data<-by(allspecies, list(allspecies$Depth, allspecies$Year), 
function(allspecies)weighted.mean(allspecies$CPUE))
but this doesn't seem to give the proper values, as the CPUE values are
included. I also have many NA's. When i try to exclude those all my values
change.

Thanks for your help!
much appreciated
marjolein

--
View this message in context:
http://r.789695.n4.nabble.com/weighted-averages-for-two-variables-tp4630589.html
Sent from the R help mailing list archive at Nabble.com.

Hello,

Post a data example, it will give us something to work on, and give you 
better answers.

dput( head(allspecies, 20) )

Then copy the output of this and paste it in your post.
As for your problem, considering it's description, and untested (no 
data), maybe


result <- by(allspecies, allspecies$Year, function(x) 
weighted.mean(x$Depth, x$CPUE))


Also, note that I've renamed the result, 'data' is an R function.

Hope this helps,

Rui Barradas

This is the output that i got
> dput( head(allspecies, 20) ) 
structure(list(Year = c(1980L, 1980L, 1980L, 1980L, 1980L, 1980L, 
1980L, 1980L, 1980L, 1980L, 1980L, 1980L, 1981L, 1981L, 1981L, 
1981L, 1981L, 1981L, 1981L, 1981L), Quarter = c(1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L
), latitude = c(54.9, 54.4833, 54.4833, 53.9167, 53.9167, 54.7833, 
58.8667, 58.25, 58.25, 57.9, 57.9, 57.9667, 56.5333, 56.5333, 
56.5333, 56.5333, 54.7167, 54.7167, 54.7167, 54.45), longitude = c(6.1333, 
7.55, 7.55, 0.9, 0.9, 1.0333, -2.3833, -3.15, -3.15, -3.3667, 
-3.3667, -2.8333, -2.2, -2.2, -2.2, -2.2, 7.3333, 7.3333, 7.3333, 
7.55), Depth = c(33, 24, 24, 49, 49, 55, 76, 69, 69, 57, 57, 
73, 46, 46, 46, 46, 20, 20, 20, 20), Area = c(6L, 6L, 6L, 4L, 
4L, 4L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 6L, 6L, 6L, 6L
), Species = structure(c(1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L), .Label = c("Agonus
cataphractus", 
"Amblyraja radiata", "Anarhichas lupus", "Arnoglossus
laterna",
"Buglossidium luteum", "Echiichthys vipera", "Eutrigla
gurnardus",
"Gadus morhua", "Glyptocephalus cynoglossus",
"Hippoglossoides
platessoides", 
"Lepidorhombus whiffiagonis", "Limanda limanda",
"Lophius piscatorius",
"Melanogrammus aeglefinus", "Merlangius merlangus",
"Merluccius merluccius",
"Microstomus kitt", "Molva molva", "Pleuronectes
platessa", "Pollachius
virens", 
"Scyliorhinus canicula", "Solea solea", "Squalus
acanthias",
"Trisopterus esmarkii", "Trisopterus luscus",
"Trisopterus minutus"
), class = "factor"), CPUE = c(1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 
1, 1, 1, 1, 1, 1, 1, 1, 1, 1)), .Names = c("Year",
"Quarter",
"latitude", "longitude", "Depth",
"Area", "Species", "CPUE"), row.names 15:34, class
= "data.frame")



--
View this message in context:
http://r.789695.n4.nabble.com/weighted-averages-for-two-variables-tp4630589p4630604.html
Sent from the R help mailing list archive at Nabble.com.

Hello,

Thanks.
The revised 'by' I had posted works with me.
An alternative, more complicated, is using split/sapply.

sapply(split(allspecies, allspecies$Year), function(x)
         weighted.mean(x$Depth, x$CPUE))


Rui Barradas

It works now :)
thank you!

--
View this message in context:
http://r.789695.n4.nabble.com/weighted-averages-for-two-variables-tp4630589p4630610.html
Sent from the R help mailing list archive at Nabble.com.