Next simulation features
Features of next simulation implied by currently selected inputs.
This is a web app for close-kin capture-recapture study power analysis by simulation. It simulates datasets from capture-recapture studies of populations of animals over time, checks whether the simulations match certain predictions, fits close-kin and/or conventional capture- recapture models to the datasets, and analyses the resulting estimates. It is most applicable to 'whale-like' populations that have one or fewer offspring per year
Features of next simulation implied by currently selected inputs.
Features of current simulation, selected and implied.
Sample-histories from first population and study simulated.
The data for the first few animals sampled in the study, by earliest birth year.
Sufficient statistics for sample histories, used to fit popan models.
Numbers of kin-pairs simulated in the first population and sampling study simulated.
Numbers of pairs of individuals with given relationships, where both are sampled in the given survey-year, or one individual is sampled in each of the given pair of survey-years. Total numbers sampled are included for reference.
Numbers of pairs of individuals with given relationships, where both are sampled in the given survey-year, or one individual is sampled in each of the given pair of survey-years. Total numbers sampled are included for reference.
Numbers of pairs of individuals with given relationships, where both are alive in the given survey-year, or one individual is alive in each of the given pair of survey-years. Population sizes are included for reference.
Genetic analysis of samples from first population and study simulated.
First few individuals sampled.
Relative frequencies, excluding samples from same animals in different surveys.
Probabilities of possible genotypes assuming that genes are inherited with probabilities given by their frequencies above.
Probabilities of all possible genopairs given each kinship.
Log-probabilities of all possible genopairs given each kinship.
Observed values for first few sample-pairs, over all loci.
Observed values for all sample-pairs, over all loci.
Possibly adjusted by a large positive factor to avoid underflow.
Observed values for first few sample-pairs, over all loci.
Numbers of pairs with corresponding survey indices for first and second samples.
Observed values for first few sample-pairs, over all loci (order is random to avoid bias due to age representation when individuals repeated to include pairs between surveys with different numbers of samples).
Numbers of pairs with corresponding survey indices for first and second samples.
Observed values for all sample-pairs, over all loci.
Negative log-likelihood surface over each parameter while others held at true values.
Parameter and kinpair probability estimates for first simulated study. Close-kin models include self, parent-offspring, and half-sibling pairs.
Numbers of individuals that are alive in the population.
The first generation is simulated with unknown parents. These individuals affect the observed numbers of kin-pairs, as parent-offspring pairs among them, and sibling-pairs including them, are both unknown. This causes the appearance of prediction error, so it should be taken into account when evaluating predictor performance.
Below are the average percentages of individuals with unknown parents, that are alive in survey-years, and pairs of survey-years. The observed numbers of pairs of individuals, including one or more with unknown parents, will be added soon.
Total numbers of pairs of individuals.
Pairs in which both individuals are alive in the same survey-year.
Pairs in which one individual is alive in each of two different survey-years.
Numbers of pairs of individuals that are the same individual alive in two different survey-years.
These include individuals with unknown parents (see unknown parents tab)
These are excluded when counting sibling-pairs from different survey-years (see unknown parents tab).
Numbers of pairs of individuals that are parent and offspring.
Pairs in which both individuals are alive in the same survey-year.
Pairs in which one individual is alive in each of two different survey-years.
Numbers of pairs of individuals with the same mother.
Pairs in which both individuals are alive in the same survey-year.
Pairs in which one individual is alive in each of two different survey-years.
Numbers of pairs of individuals with the same father.
Pairs in which both individuals are alive in the same survey-year.
Pairs in which one individual is alive in each of two different survey-years.
Numbers of pairs of individuals with the same parents.
Pairs in which both individuals are alive in the same survey-year.
Pairs in which one individual is alive in each of two different survey-years.
Numbers of pairs of individuals that share exactly one parent.
Pairs in which both individuals are alive in the same survey-year.
Pairs in which one individual is alive in each of two different survey-years.
Average proportional differences between numbers simulated and predicted, over survey-years and pairs of survey-years. Some of these differences are affected by individuals with unknown parents (see unknown parents tab), so these are also reported below. The prediction for the total number of pairs of individuals (all pairs) seems to show a consistent difference. The number of full-sibling pairs is much smaller than the others so the proportional difference from the prediction seems to be more variable.
Average percentages of individuals with unknown parents, that are alive in survey-years, and pairs of survey-years.
Numbers of pairs of individuals with given relationships, where both are alive in the same survey-year. Population sizes are included for reference.
Numbers of pairs of individuals with given relationships, where one individual is alive in each of a pair of survey-years.