Prior to analyzing the data, this function initializes several variables (e.g., number of capture occasions, time intervals) that are often specific to the capture-recapture model being fitted to the data. It also is used to 1) define groups in the data that represent different levels of one or morestrata.labels factor covariates (e.g., sex), 2) define time intervals between capture occasions (if not 1), and 3) create an age structure for the data, if any.
process.data(data,begin.time=1,model="CJS",mixtures=1,groups=NULL,
allgroups=FALSE,age.var=NULL,initial.ages=c(0),
time.intervals=NULL,nocc=NULL,accumulate=TRUE,
strata.labels=NULL) accumulate_data(data)
from process.data processed.data (a list with the following elements)
original raw dataframe with group factor variable added if groups were defined
type of analysis model (eg, "cjs" or "js")
a dataframe of frequencies (same # of rows as data, number of columns is the number of groups in the data. The column names are the group labels representing the unique groups that have one or more capture histories.
number of capture occasions
length of time intervals between capture occasions
time of first capture occasion
an initial age for each group in the data; Note that this is not the original argument but is a vector with the initial age for each group.
factor covariates used to define groups
from accumulate_data a dataframe with same column structure as argument with addition of freq (if not any) and reduced to unique rows with freq accumulating number of records.
A data frame with at least one field named ch which is
the capture (encounter) history stored as a character string. data
can also have a field freq which is the number of animals with that
capture history. The default structure is freq=1 and it need not be included
in the dataframe. data can also contain an arbitrary number of
covariates specific to animals with that capture history.
Time of first capture occasion or vector of times if different for each group
Type of analysis model.
Number of mixtures in closed capture models with heterogeneity
Vector of factor variable names (in double quotes) in
data that will be used to create groups in the data. A group is
created for each unique combination of the levels of the factor variables in
the list.
Logical variable; if TRUE, all groups are created from
factors defined in groups even if there are no observations in the
group
An index in vector groups for a variable (if any) for
age
A vector of initial ages that contains a value for each
level of the age variable groups[age.var]; if the data contain an initial.age field then it will be used instead.
Vector of lengths of time between capture occasions or matrix of time intervals with a row for each animal and column for each interval between occasions.
number of occasions for Nest type; either nocc or time.intervals must be specified
if TRUE, aggregates data with same values and creates freq field for count of records
labels for strata used in capture history; they are converted to numeric in the order listed. Only needed to specify unobserved strata; for any unobserved strata p=0.
Jeff Laake
For examples of data, see dipper. The structure of the
encounter history and the analysis depends on the analysis model to some
extent. Thus, it is necessary to process a dataframe with the encounter
history (ch) and a chosen model to define the relevant values.
For example, number of capture occasions (nocc) is automatically
computed based on the length of the encounter history (ch) in
data.
The default time interval is unit time (1) and if this is
adequate, the function will assign the appropriate length. A processed data
frame can only be analyzed using the model that was specified. The
model value is used by the functions make.design.data
and crm to define the model structure as it relates to the
data. Thus, if the data are going to be analysed with different underlying
models, create different processed data sets with the model name as an
extension. For example, dipper.cjs=process.data(dipper).
This function will report inconsistencies in the lengths of the capture history values and when invalid entries are given in the capture history.
The argument begin.time specifies the time for the first capture
occasion and not the first time the particular animal was caught or releaed. This is used in creating the levels of the time factor variable
in the design data and for labelling parameters. If the begin.time
varies by group, enter a vector of times with one for each group. It will add a field
begin.time to the data with the value for each individual. You can also specify a
begin.time field in the data allowing each animal to have a unique begin.time. Note that
the time values for survivals are based on the beginning of the survival
interval and capture probabilities are labeled based on the time of the
capture occasion. Likewise, age labels for survival are the ages at the
beginning times of the intervals and for capture probabilities it is the age
at the time of capture/recapture.
The time.intervals argument can either be a vector of lengths of times for each interval between occasions that is constant for all animals or a matrix which has a row for each animal and a column for each interval which lets the intervals vary by animals. These intervals are used to construct the design data and are used for the field time.interval which is used to adjust parameters like Phi and S to a constant per unit time interval (eg annual survival rates). On occasion it can be useful to leave the time.interval to remain at default of 1 or some other vector of time.intervals to construct the design data and then modify the time.interval value in the design data. For example, assume that cohort marking and release is done between sampling occasions. The initial survival from release to the next sampling occasion may vary by release cohort, but the remainder of the surivivals are between sampling occasions. In that case it is easier to let time.interval=1 (assuming unit interval (eg year) between sampling occasions but then modifying ddl$Phi$time.interval to the value for the first interval after each release to be the partial year from release to next sampling occasion. In this way everything is labelled with annual quantities but the first partial year survival is adjusted to an annual rate.
Note that if you specify time.intervals as a matrix, then accumulate is set to FALSE so that the number of rows in the data can be checked against the number of rows in the time.intervals matrix and thus data cannot be accumulated because at present it doesn't use values of time.intervals to determine which records can be accumulated.
groups is a vector of variable names that are contained in
data. Each must be a factor variable. A group is created for each
unique combination of the levels of the factor variables. For
example groups=c("sex","ageclass","region") creates
groups defined by the levels of sex, ageclass and region.
the code will only use groups that have 1 or more capture histories unless allgroups=TRUE.
The argument age.var=2 specifies that the second grouping variable in
groups represents an age variable. It could have been named
something different than ageclass but it should not be named age as that is reserved in marked.
initial.age specifies that the age at first capture of the age levels. For example
initial.age=0:2 specifies that the initial.ages are 0,1 and 2 for the age class levels
designated as 1,2,3. The actual ages
for the age classes do not have to be sequential or ordered, but ordering
will cause less confusion. Thus levels 1,2,3 could represent initial ages
of 0,4,6 or 6,0,4. The default for initial.age
is 0 for each group, in which case, age represents time since marking
(first capture) rather than the actual age of the animal. If the data contains an initial.age field
then it overrides any other values and lets each animal have a unique initial.age at first capture/release.
The following variable names are reserved and should not be used in the data: id (animal id) ch(capture history) freq (number of animals with that ch/data) occ,age,time,cohort,Age,Time,Cohort,time.interval,fix
dipper,crm
data(dipper)
dipper.process=process.data(dipper,groups="sex")
# create some artificial age data as an example
dipper$ageclass=factor(c(rep("A",100),rep("J",194)))
dipper.process=process.data(dipper,groups=c("sex","ageclass"),age.var=2,initial.ages=c(1,0))
Run the code above in your browser using DataLab