Some studies are intended to compare regular trajectories of the same
duration collected at different period. For example, the aim may be to
identify the differences/similarities between different days (each one
corresponding to a burst of relocation) in the pattern of movements of
an animal between 05H00 and 08H00, with a time lag of 5 minutes. In
such cases, it is often convenient that the relocations of the bursts
are paired (e.g. the fifth relocation correspond to the position of
the animal at 5H30 for all bursts).
The function set.limits
is intended to ensure that the time of
beginning, the end, and the duration of the trajectory is the same for
all bursts of the object ltraj
. If relocations are collected
outside the limits, they are removed. If the actual time limits of
the burst cover a shorter period than those specified, missing values
are added.
Note that "time of beginning" is not a synonym for "date". That is,
two trajectories of the same animal, both beginning at 05H00 and
ending at 08H00, have the same time of beginning, but are necessarily
not sampled on the same day, which implies that they correspond to
different dates. For this reason, the time of beginning is indicated
to the function set.limits
by a character string, and the
parameter pattern
should indicate the conversion specifications.
These conversions specifications are widely documented on the help
page of the function strptime
. For example, to indicate that the
trajectory begins at 5H00, the value for begin
should be
"05:00"
and the value for pattern
should be
"%H:%M"
. If the trajectory should begin on january 10th, the
value for begin
should be "01:10"
and pattern
should be "%m:%d"
. Note that the only conversion
specifications allowed in this function are %S
(seconds),
%M
(minutes), %H
(hours), %d
(day), %m
(month), %Y
(year with century), %w
(weekday), and
%j
(yearday). See help(strptime)
for additional
information on these convention specifications.