Dear All,
I am interested in using R to summarize data so that I can THEN do some
additional analyses using behavioral data.
I study birds, mostly, and often behavioral data that includes timing. So,
for example, each line may be the occurrence of some behavior, and it would
include identifiers for species, individual, date, time of observation and
the various behavioral observations that may be important. Thus, if
observing one individual, the difference in the time of one line to the
subsequent line could be the time it took to carry out a behavior. Or,
perhaps several lines include all the behaviors of a "bout" of
behaviors.
For example, I have a project on hummingbirds, in which an individual can be
observed for a time interval, during which it may do several things. So,
each "thing" becomes a line to identify what it was. Then, I would
like to
count behaviors from the start to the end of his "bout." Basically, I
want
to calculate a few things - the difference between the times of the
behaviors of one line and the subsequent or some number of subsequent lines.
I want to count the number of other things in a time interval, and so on.
In JMP, to calculate the interval from one line to the next, I would
generate a new variable, and that basically subtracts the time in line x
from the time in line x+1. I can't figure out how to do that in R.
Next, in JMP (or SAS), I would do a summary, in which I based it on date,
species, individual, etc., and which took a Maximum and Minimum for time,
the sum of say variable X, the Maximum of variable Y and so on.
Example:
Date Begin End Species Individual Feeding
Interaction Interactsp
2008-09-11 11:15:21 11:15:58 CHAU 1
3
2008-09-11 11:16:01 11:16:12 CHAU 1
Chase CHAU
2008-09-11 11:16:17 11:16:28 CHAU 1 5
2008-09-11 11:16:32 11:16:49 CHAU 1
Flee THGL
2008-09-11 11:16:53 11:17:15 THGL 1 8
Clearly, to make another variably by just subtracting one variable from
another, like time begin time end, is no problem. Also, since the species
changed, the individual number can remain. What I would like to do, for
example, is count the number of feedings, interactions, and so on in an
interval, as well as the number of species, and calculate the total time the
individual was at the food source.
Well, all that is very simple in JMP and I just can't figure out how to do
it in R. In fact, just figuring out how to make time and data TIME and DATE
was a bit complex, but I have managed to do that. I hope I have explained my
problem well enough, because all the other problems I have hinge upon
summarizing my data in a way that the summary can then be used in analysis.
If I could figure out this one part, I would probably switch to R
completely!
If you have any helpful suggestions, I am sure that there are other
behavioral biologists out there that would be able to benefit from this kind
of information. I know my future students will!
Sincerely,
Jim
--
James J. Roper
Smithsonian Tropical Research Institute
Bocas del Toro Marine Research Station
MRC 0580-03
Unit 9100, Box 0948
DPO AA 34002-9998
Skype-in (USA):+1 706 5501064
Skype-in (Brazil): 41 39415715
E-mail - personal: jjroper@gmail.com
E-mail - consulting: arsartium@gmail.com
9 21.122' N, and 82 15.390' W
In Google Earth, copy and paste -> 9 21.122' N, 82 15.390' W
Ecologia e Conservação na UFPR
http://www.bio.ufpr.br/ecologia/
Personal Pages
http://jjroper.googlespages.com
http://arsartium.googlepages.com
[[alternative HTML version deleted]]
