CODEWITHSUNDEEP
raggregateaverage
striatum
striatum

Reputation: 1598

Can `ave()` in `R` aggregate more than one variable over the list of variables as grouping elements

THIS POST seems relevant, but the example considers only one grouping variable (by). Is it possible to use ave() for several variables to be averaged on the basis of several grouping variables?

Following the example in the same POST:

df <- data.frame(A = rnorm(20), B = rnorm(20), site = gl(5,4), month = gl(2,10))

this seems to work fine, still using mean not ave

aggregate(cbind(A, B) ~ site + month, data = df, mean)

Is there a more efficient way? Can ave() do that straight away?

Upvotes: 2

Views: 72

Answers (1)

jay.sf
jay.sf

Reputation: 73572

You can add more grouping variables.

> transform(df, mn=ave(cbind(A, B), site, month))
            A          B site month        mn.A        mn.B
1  -0.6104697  0.2524263    1     1 -0.03415871 -0.03415871
2   0.2880355  0.1897356    1     1 -0.03415871 -0.03415871
3  -0.5816155  1.2815789    1     1 -0.03415871 -0.03415871
4   1.3435948 -2.4365555    1     1 -0.03415871 -0.03415871
5   0.1750469 -0.5855076    2     1 -0.10553138 -0.10553138
6   0.4374662  0.0103783    2     1 -0.10553138 -0.10553138
7  -0.3405101 -0.8371708    2     1 -0.10553138 -0.10553138
8   0.6288837 -0.3328376    2     1 -0.10553138 -0.10553138
9  -1.6613419 -2.1050023    3     1 -0.72058797 -0.72058797
10  1.0453300 -0.1613377    3     1 -0.72058797 -0.72058797
11  0.5053064 -1.2946748    3     2 -0.13016572 -0.13016572
12 -1.9228812  2.1915868    3     2 -0.13016572 -0.13016572
13 -1.4860137 -0.3145034    4     2 -0.66294349 -0.66294349
14 -0.8653919 -0.3814499    4     2 -0.66294349 -0.66294349
15  0.1556521 -1.1469195    4     2 -0.66294349 -0.66294349
16 -0.9116240 -0.3532977    4     2 -0.66294349 -0.66294349
17  1.1310613  2.0196141    5     2  0.63761020  0.63761020
18 -0.6198639  0.8882268    5     2  0.63761020  0.63761020
19  0.3932094 -1.3104097    5     2  0.63761020  0.63761020
20  1.6947693  0.9042743    5     2  0.63761020  0.63761020

Note, that this is actually ave(cbind(A, B), site, month, FUN=mean), i.e. defaulting to mean.

Upvotes: 2

Related Questions