cooccurence_test.Rd
Maximum likelihood test and plotted results for coinfection dynamics
cooccurence_test( data, density_func = independent, max_moi = 25, poisson = FALSE, size = 100, boot_iter = 10000, plot = TRUE, quantiles = c(0.025, 0.975), lower = NULL, upper = NULL, ... )
data | Observed data. Either as a named vector or a 2 column
|
---|---|
density_func | Function that calculates the multinomial distribution
describing the data. Default = |
max_moi | Maxuimum infction composition explored. Default = 25. |
poisson | Logical for determining if we use a poisson distribution to describe the numer of infection occurences. Default = `FALSE`, which means a negative binomial is used. ` |
size | Starting value for the negative binomial size parameter. Default = 100. |
boot_iter | Bootstrap iterations. Default = 10000 |
plot | Boolean for default plotting the bootsrap results. Default = TRUE |
quantiles | Vector of length 2 for the quantiles used. Default = `c(0.025, 0.975)` |
lower | Vector of lower bounds used in fitting. Default = NULL, which will create a vector with 0.0001 for each frequency and 0.1 for the moi. |
upper | Vector of upper bounds used in fitting. Default = NULL, which will create a vector with 0.9999 for each frequency and `max_moi` for the moi. |
... | Any other arguments that will be passed to `density_func`` |
Invisibly returns a list containing the estimated parameters and a plot of our data compared to the boostrapped estimates.
Estimates the maximum likely population frequency of each species (i.e. the names of the observed entities in our data) and the mean number of infections given our observed data. These estimates are used to estimate the multinomial probability distribution for all cooccurences and comparing these to our data using a boostrap method.
if (FALSE) { # example of the two forms of data type accepted real <- data.frame( "variable"=c("pf/po","pf/pv","pf/po/pv","pf","po","po/pv","pv"), "value"=c(84,179,1,5181,44,1,309) ) real <- c( "pf/po" = 84, "pf/pv" = 179, "pf/po/pv" = 1, "pf" = 5181, "po" = 44, "po/pv" = 1, "pv" = 309 ) res <- cooccurence_test(data = real) }